# Admin
Source: https://docs.parallellabs.app/account/admin

Platform-wide administration: usage, costs, users, companies, and feature flags.

The **Admin** area is for operating the platform itself rather than working inside one company. It sits at the bottom of the sidebar and covers every company and user on the instance.

> **Super users only.** Admin is visible only to super users, and the page itself refuses to render for anyone else. Super user status is set directly in the database — there is deliberately no way to grant or revoke it through the interface.

Unlike everything else in the sidebar, Admin is not scoped to a company.

## Sections

| Section            | What it covers                                                                                                                                              |
| ------------------ | ----------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Analytics**      | Platform usage — companies, users, logins, and breakdowns by feature, model, and method, with credits, errors, average duration, and last-month comparisons |
| **Costs**          | Cost breakdowns by company, organization, user, feature, and model, including input and output token detail                                                 |
| **Users**          | Every user, with subscription, credits, limits, created date, and last login; display name, data credits, credit limit, and company limit are editable      |
| **Companies**      | Every company, with owner, members, and credits; edit the name, credit limit, and whether MFA is required, and manage members                               |
| **Terminations**   | Companies scheduled for termination, with reason, scheduled date, days until cutoff, and status — and the option to cancel one                              |
| **Audiences**      | Saved audiences across companies, with refresh interval, last refresh, and notification settings, plus a manual refresh                                     |
| **Pixels**         | Tracking pixels, with their website, company, list, tags, and status                                                                                        |
| **Mailbox Orders** | Mailbox provisioning orders, with domains, mailbox counts, totals, and status; orders can be cancelled                                                      |
| **Notifications**  | Platform notifications — announcements, maintenance notices, new feature notes, info, and errors — filterable by type and scope                             |
| **Feature Flags**  | Feature availability, enabled for everyone or targeted at specific companies or users                                                                       |

**Analytics** is the default view.

## Common Tasks

**Adjusting a user's limits.** Open **Users**, find the user, and edit their data credits, credit limit, or company limit directly in the row.

**Investigating cost.** **Costs** breaks spend down by company, user, feature, and model — the fastest way to find what's driving an unexpected bill.

**Rolling out a feature gradually.** In **Feature Flags**, target specific companies or users first, then enable for everyone once it holds up.

**Stopping a scheduled deletion.** **Terminations** lists companies heading for cutoff, with the days remaining and a cancel action.

## Cautions

1. **Changes here affect real customers immediately.** There's no staging copy of this data.
2. **Credit and limit edits take effect at once**, including mid-billing-period.
3. **Feature flags are the safer lever.** Prefer targeting a few accounts over enabling something for everyone.
4. **Terminations are time-sensitive.** Once past cutoff, cancelling is no longer the remedy.

## Next Steps

* [Manage a single company](/account/companies)
* [Manage team access](/settings/collaborators)


# Companies
Source: https://docs.parallellabs.app/account/companies

Create, switch between, and manage the companies that hold your work.

Companies (also called organizations) are your workspaces in the platform. Each company has its own AI employees, knowledge base, campaigns, and settings. You might create separate companies for different businesses, clients, or departments.

## Creating Your First Company

1. Click **Companies** in the sidebar (or you'll be prompted when you first log in)
2. Click the **Create Company** button
3. Enter your company name
4. (Optional) Upload a company logo
5. Click **Create**

That's it! You now have a workspace ready to use.

## Switching Between Companies

If you have multiple companies:

1. Click the company name dropdown at the top of the sidebar
2. Select the company you want to work in
3. The page will refresh with that company's data

**Tip:** You can also access all your companies by clicking **Companies** in the sidebar menu.

## Editing a Company

To change a company's name or logo:

1. Go to **Companies** in the sidebar
2. Find the company you want to edit
3. Click the **three-dot menu** (⋮) on the right side of that row
4. Select **Edit**
5. Make your changes
6. Click **Save**

## Reordering Your Companies

Want to change the order companies appear in your dropdown? You can drag and drop them:

1. Go to **Companies** in the sidebar
2. Hover over a company row - you'll see a drag handle on the left
3. Click and drag the company to your preferred position
4. Release to drop it in place

The new order is saved automatically.

## Deleting a Company

**Warning:** Deleting a company permanently removes all its data, including AI employees, documents, sequences, and conversations. This cannot be undone.

To delete a company:

1. Go to **Companies** in the sidebar
2. Click the **three-dot menu** (⋮) next to the company
3. Select **Delete**
4. Confirm by clicking **Delete** in the popup

## Company Limits

Your account has a limit on how many companies you can create:

* **Free and Entrepreneur plans:** 1 company
* **Business plan:** 3 companies included
* **Enterprise:** Custom limits

Paid plans can typically add more companies for an additional monthly fee, and your limit scales with the number of seats on your subscription.

To see your current limits:

1. Go to **Profile** in the sidebar
2. Look for "Companies" in your usage section

Need more companies? Upgrade your subscription or contact support.

## Transferring Company Ownership

If you need to transfer a company to another user (for example, handing off a client's workspace):

1. Go to **Companies** in the sidebar
2. Click the **three-dot menu** (⋮) next to the company
3. Select **Transfer Ownership**
4. Enter the email address of the new owner
5. Click **Transfer**

The new owner will receive the company in their account. You will lose owner access but may retain access as a collaborator depending on settings.

## Frequently Asked Questions

**Q: Can I be part of companies I don't own?**
A: Yes! Other users can invite you as a collaborator to their companies. You'll see these in your company list but won't be able to delete them.

**Q: What's the difference between a company and a team?**
A: A company is your workspace. Team members (collaborators) are people you invite to work with you inside that company.

**Q: Can I copy data from one company to another?**
A: Currently, data must be manually recreated in each company. We recommend exporting any lists or documents you need before deleting a company.

## Next Steps

Now that you have a company set up:

* [Create your first AI Employee](/overview/employees)
* [Add documents to your Knowledge Base](/overview/knowledge-base)
* [Invite team members](/settings/collaborators)


# Account vs. Company
Source: https://docs.parallellabs.app/account/overview

Settings that follow you across every company you have access to.

Your login, subscription, credit balance, and security settings belong to you, not to any one company. Switching companies doesn't change them.

## In This Section

* **[Companies](/account/companies)** — Create, switch between, reorder, and manage your companies.
* **[Profile](/account/profile)** — Your personal details, subscription, credits, notifications, and security settings.
* **[Support](/account/support)** — Submit and track support tickets.

## Companies vs. Your Account

A **company** is a workspace: its own AI employees, lists, content, integrations, and collaborators. Your **account** is you — one login that can hold several companies.

That's why credits and subscription live on your [Profile](/account/profile) rather than inside a company, while integrations and collaborators are set per company.

How many companies you can create depends on your plan — see [Companies](/account/companies).


# Profile
Source: https://docs.parallellabs.app/account/profile

Manage your personal details, subscription, credits, notifications, and security.

Your profile page is where you manage your personal account settings, view your subscription, and track your usage across the platform.

## Accessing Your Profile

Click **Profile** in the sidebar menu (near the bottom).

## Updating Your Display Name

Your display name appears in conversations and when collaborating with team members.

To change it:

1. Go to **Profile**
2. Click on your current name (or the edit icon next to it)
3. Type your new name
4. Click **Save**

## Changing Your Profile Photo

1. Go to **Profile**
2. Click on your current photo (or the camera icon)
3. Select an image from your computer
4. The photo uploads automatically

**Tip:** Use a square image for best results. The photo will be cropped to fit a circle.

## Understanding Your Subscription

Your profile shows your current subscription plan and what's included:

### What You'll See

| Item             | What It Means                                                        |
| ---------------- | -------------------------------------------------------------------- |
| **Current Plan** | Your subscription tier (Free, Entrepreneur, Business, or Enterprise) |
| **Credit Usage** | How much of your monthly AI credits you've used                      |
| **Credit Limit** | Your total monthly allowance                                         |
| **Seats**        | How many team members can use your account                           |
| **Companies**    | How many workspaces you can create                                   |

### The Usage Bar

The progress bar shows your credit consumption:

* **Green** = You have plenty of credits remaining
* **Yellow** = You're approaching your limit
* **Red** = You're at or over your limit

Credits reset at the beginning of each billing cycle.

## Upgrading Your Subscription

Need more credits, seats, or features? Here's how to upgrade:

1. Go to **Profile**
2. Click **Upgrade** or **Manage Subscription**
3. Choose a new plan
4. Enter your payment information (if not already on file)
5. Confirm your upgrade

Your new features are available immediately. You'll be charged the prorated difference for the current billing period.

## Managing Payment Methods

To update your credit card or payment method:

1. Go to **Profile**
2. Click **Manage Subscription** or **Billing**
3. You'll be taken to our secure payment portal
4. Update your payment method
5. Save changes

## What Are Credits?

Credits are the currency that powers AI features in the platform. Different actions use different amounts of credits:

| Action              | Credit Cost |
| ------------------- | ----------- |
| Chat messages       | Low         |
| Document processing | Medium      |
| Image generation    | Medium-High |
| Video generation    | High        |
| Audio generation    | Medium      |

Your plan includes a monthly credit allowance. If you run out, you can:

* Wait for credits to reset next month
* Upgrade to a higher plan
* Purchase additional credits (if available on your plan)

## Account Settings

### Email Notifications

Control what emails you receive:

1. Go to **Profile**
2. Scroll to **Notifications** or **Settings**
3. Toggle notifications on/off for different types

### Dark Mode

Prefer a darker interface? The light/dark toggle lives in the sidebar (look for the moon/sun icon), not on the Profile page. Click it to switch between light and dark modes at any time.

## Security

### Your Password

You sign in with a magic link, a verification code, or Google — so there's no password to manage day to day. If you use an email-and-password login and need to reset it, use the **Forgot password?** link on the sign-in screen; a reset link will be emailed to you.

### Two-Factor Authentication

For extra security, enable two-factor authentication (2FA):

1. Go to **Profile**
2. Find the **Security** or **Two-Factor Authentication** section
3. Follow the setup steps to link your authenticator app
4. Enter the 6-digit code from your app to confirm

Once enabled, you'll enter a 6-digit code from your authenticator app each time you sign in. You can disable 2FA from the same section (you'll be asked to confirm with a code).

## Frequently Asked Questions

**Q: What happens if I run out of credits?**
A: Most AI-powered features will be temporarily unavailable until your credits reset or you upgrade. You can still access your data and non-AI features.

**Q: Can I downgrade my subscription?**
A: Yes. Go to Profile → Manage Subscription and select a lower tier. Changes take effect at your next billing cycle.

**Q: How do I cancel my subscription?**
A: Go to Profile → Manage Subscription → Cancel. You'll retain access until the end of your current billing period.

**Q: I was charged incorrectly. What do I do?**
A: Contact support through the Support page. Include your billing date and the charge amount.

## Next Steps

* [Create a Company](/account/companies)
* [Get Support](/account/support)
* [Invite Team Members](/settings/collaborators)


# Support
Source: https://docs.parallellabs.app/account/support

Submit and track support tickets, and find help when you need it.

Need help? We're here for you! The Support page lets you submit questions, report issues, and track your support requests.

> Depending on your account, clicking **Support** may open the in-app ticket form described here or take you to your provider's dedicated support channel (help desk, email, or community). Either way, it's the fastest path to reach the right team.

## Submitting a Support Ticket

1. Click **Support** in the sidebar
2. Click **Create Ticket** (or the **+** button)
3. Fill out the form:
   * **Subject:** A brief description of your issue
   * **Category:** Select what type of help you need
   * **Description:** Explain your issue in detail
   * **Attachments:** Add screenshots or files if helpful
4. Click **Submit**

You'll receive a confirmation, and our team will respond as soon as possible.

## Writing a Good Support Request

The more details you provide, the faster we can help. Include:

**For bug reports:**

* What you were trying to do
* What happened instead
* Steps to reproduce the issue
* Screenshots or screen recordings
* Which browser you're using

**For questions:**

* What you're trying to accomplish
* What you've already tried
* Any error messages you've seen

**Example of a good ticket:**

> **Subject:** Can't upload PDF to Knowledge Base
>
> **Description:** I'm trying to upload a PDF document to my Knowledge Base but it fails every time. Here's what happens:
>
> 1. I click "Upload Document"
> 2. I select my PDF file (attached)
> 3. The progress bar reaches 100%
> 4. Then I see "Upload failed" error
>
> The file is 2.5MB. I'm using Chrome on Mac.
>
> \[screenshot attached]

## Choosing the Right Category

| Category            | When to Use                                     |
| ------------------- | ----------------------------------------------- |
| **Bug Report**      | Something isn't working correctly               |
| **Feature Request** | You'd like us to add something new              |
| **Account**         | Login issues, password problems, account access |
| **Billing**         | Questions about charges, subscriptions, refunds |
| **Integration**     | Problems connecting to other tools              |
| **General**         | Anything else                                   |

## Checking Your Ticket Status

To see updates on your tickets:

1. Go to **Support** in the sidebar
2. Your open tickets appear in the list
3. Click on a ticket to see the full conversation

### Ticket Statuses

| Status             | What It Means                           |
| ------------------ | --------------------------------------- |
| **Open**           | We've received it and will respond soon |
| **In Progress**    | We're actively working on it            |
| **Waiting on You** | We need more information from you       |
| **Resolved**       | The issue has been addressed            |
| **Closed**         | The ticket is complete                  |

## Responding to Support

When we reply to your ticket:

1. You'll receive an email notification
2. Click the link in the email, or go to Support in the app
3. Click on your ticket
4. Type your response in the reply box
5. Click **Send**

## Searching Your Tickets

Looking for an old conversation?

1. Go to **Support**
2. Use the search bar at the top
3. Type keywords from your issue
4. Matching tickets will appear

## Response Times

We aim to respond to all tickets within:

| Priority                         | Response Time   |
| -------------------------------- | --------------- |
| Critical (can't use platform)    | Within hours    |
| High (feature broken)            | Within 24 hours |
| Normal (questions, minor issues) | Within 48 hours |
| Feature requests                 | Reviewed weekly |

**Tip:** If your issue is urgent, mention it in the subject line (e.g., "URGENT: Can't access my account").

## Self-Service Help

Before submitting a ticket, you might find your answer in:

* **This documentation** - Browse the guides for your feature
* **In-app tooltips** - Hover over (?) icons for quick help
* **FAQ sections** - Check the bottom of each documentation page

## Common Issues & Quick Fixes

### "I can't log in"

* Make sure you're using the correct email
* Try resetting your password
* Clear your browser cache and cookies
* Try a different browser

### "My changes aren't saving"

* Check your internet connection
* Refresh the page and try again
* Make sure you clicked the Save button (some forms require this)

### "A feature is missing"

* Check if the feature is locked (requires upgrade)
* Make sure you have the right permissions
* Verify you're in the correct company

### "I'm seeing an error message"

* Take a screenshot of the error
* Note what you were doing when it appeared
* Submit a ticket with these details

## Feedback & Feature Requests

We love hearing your ideas! To suggest a new feature:

1. Go to **Support**
2. Create a ticket with category **Feature Request**
3. Describe what you'd like to see and why it would help you
4. We review all suggestions and prioritize based on demand

## Frequently Asked Questions

**Q: Can I call support?**
A: We currently offer support through tickets only. This helps us track issues and provide thorough responses.

**Q: How do I escalate an urgent issue?**
A: Include "URGENT" in your subject line and clearly explain the business impact.

**Q: Can I reopen a closed ticket?**
A: Yes, simply reply to the ticket and it will reopen automatically.

**Q: Who can submit support tickets?**
A: Any user with an account can submit tickets for issues they're experiencing.

## Next Steps

* [Manage Your Profile](/account/profile)
* [Manage Companies](/account/companies)
* [Back to the User Guide](/getting-started)


# Cheap polling probe for whether a conversation has changed.
Source: https://docs.parallellabs.app/api-reference/agents/cheap-polling-probe-for-whether-a-conversation-has-changed

/openapi.json get /api/v0/conversations/{agentId}/{sessionId}/check
Returns when the conversation last changed, without loading any message
content. Clients compare the values to what they already have and only
refetch the full conversation when they differ.



# Create a new AI agent.
Source: https://docs.parallellabs.app/api-reference/agents/create-a-new-ai-agent

/openapi.json post /api/v0/agents
Creates a new AI agent with the specified configuration. The agent can
be configured with multiple channels (website widget, email, SMS, WhatsApp, voice)
and custom branding.



# Delete an agent.
Source: https://docs.parallellabs.app/api-reference/agents/delete-an-agent

/openapi.json delete /api/v0/agents/{agentId}
Permanently deletes the specified agent and all associated data.



# Get a specific agent by ID.
Source: https://docs.parallellabs.app/api-reference/agents/get-a-specific-agent-by-id

/openapi.json get /api/v0/agents/{agentId}
Returns the full agent configuration including channel settings,
lead fields, and branding options.



# Get all conversations across company agents with pagination and filtering.
Source: https://docs.parallellabs.app/api-reference/agents/get-all-conversations-across-company-agents-with-pagination-and-filtering

/openapi.json get /api/v0/conversations
Returns a paginated list of conversations for all agents in the company.
Filter by agentId, channel, and date range using Unix timestamps.



# Get full conversation details for a specific agent session.
Source: https://docs.parallellabs.app/api-reference/agents/get-full-conversation-details-for-a-specific-agent-session

/openapi.json get /api/v0/conversations/{agentId}/{sessionId}
Returns the full conversation including all messages exchanged between
the user and the agent. Messages are ordered chronologically.



# List all agents for the current company.
Source: https://docs.parallellabs.app/api-reference/agents/list-all-agents-for-the-current-company

/openapi.json get /api/v0/agents
Returns a paginated list of agents belonging to the company associated
with the API key. Use the search parameter to filter agents by name.



# Update an existing agent.
Source: https://docs.parallellabs.app/api-reference/agents/update-an-existing-agent

/openapi.json put /api/v0/agents/{agentId}
Updates the agent configuration. Only include fields that need to be
changed - unspecified fields will retain their current values.



# Create audio from text using text-to-speech.
Source: https://docs.parallellabs.app/api-reference/audio/create-audio-from-text-using-text-to-speech

/openapi.json post /api/v0/audio
Converts the provided text to speech using ElevenLabs.
Returns the audio ID for retrieval via the /audio/p/<audioId> endpoint.



# Delete an audio file by ID.
Source: https://docs.parallellabs.app/api-reference/audio/delete-an-audio-file-by-id

/openapi.json delete /api/v0/audio/{audioId}



# Get available audio models.
Source: https://docs.parallellabs.app/api-reference/audio/get-available-audio-models

/openapi.json get /api/v0/audio/models
Returns a list of available text-to-speech models and their capabilities.
This endpoint is public and does not require authentication.



# Get available voices for audio generation.
Source: https://docs.parallellabs.app/api-reference/audio/get-available-voices-for-audio-generation

/openapi.json get /api/v0/audio/voices
Returns both public voices and company-specific cloned voices.



# List all audio files.
Source: https://docs.parallellabs.app/api-reference/audio/list-all-audio-files

/openapi.json get /api/v0/audio
Returns a paginated list of audio files for the company.



# Build a full brand kit (name, description, style, logos) from a website or an uploaded image.
Source: https://docs.parallellabs.app/api-reference/brand-kit/build-a-full-brand-kit-name-description-style-logos-from-a-website-or-an-uploaded-image

/openapi.json post /api/v0/brand-kits/build
An agent screenshots and reads the source itself to assemble the whole brand
kit. The work runs in the background: this creates a draft brand kit row
with buildStatus 'building' and returns its id. Poll GET
/brand-kits/<id> — buildStatus flips to 'ready' (fields filled) or
'failed' (buildError set). Provide exactly one of websiteUrl / imageDataUri.



# Create a new (empty) brand kit for landing pages / websites.
Source: https://docs.parallellabs.app/api-reference/brand-kit/create-a-new-empty-brand-kit-for-landing-pages-websites

/openapi.json post /api/v0/brand-kits/create



# Create or update a brand kit.
Source: https://docs.parallellabs.app/api-reference/brand-kit/create-or-update-a-brand-kit

/openapi.json post /api/v0/brand-kits



# Delete a brand kit.
Source: https://docs.parallellabs.app/api-reference/brand-kit/delete-a-brand-kit

/openapi.json delete /api/v0/brand-kits/{brandKitId}



# Generate test images from a brand kit configuration
Source: https://docs.parallellabs.app/api-reference/brand-kit/generate-test-images-from-a-brand-kit-configuration

/openapi.json post /api/v0/brand-kits/test
Runs the content-engine image pipeline (AI image or canvas agent, per the
supplied configuration) against sample text or an existing content piece
and returns the generated image URLs without attaching them to any
content. Charges the same credits as normal generation.



# Get a specific brand kit by ID.
Source: https://docs.parallellabs.app/api-reference/brand-kit/get-a-specific-brand-kit-by-id

/openapi.json get /api/v0/brand-kits/{brandKitId}



# List a company's brand kits.
Source: https://docs.parallellabs.app/api-reference/brand-kit/list-a-companys-brand-kits

/openapi.json get /api/v0/brand-kits
Default: a paginated list. With `?listAll=true`, returns the full list
(default first). With `?landingPageId=` or `?brandKitId=`, resolves the
single kit a landing page / website should use (page-specific → company
default), for the landing-page / website renderers.



# Save a brand kit's landing-page / website assets (favicon, logo, images,
Source: https://docs.parallellabs.app/api-reference/brand-kit/save-a-brand-kits-landing-page-website-assets-favicon-logo-images

/openapi.json put /api/v0/brand-kits
CSS, head scripts, guidelines), with automatic S3 upload for images.



# Upload a single brand kit image (favicon/logo/custom) via multipart form data.
Source: https://docs.parallellabs.app/api-reference/brand-kit/upload-a-single-brand-kit-image-faviconlogocustom-via-multipart-form-data

/openapi.json post /api/v0/brand-kits/upload



# Cancel a pending or running browser task.
Source: https://docs.parallellabs.app/api-reference/browser/cancel-a-pending-or-running-browser-task

/openapi.json delete /api/v0/browser-task/{task_id}
Cancels the specified browser task if it's still pending or running.
Completed or failed tasks cannot be cancelled.



# Get browser task status and result.
Source: https://docs.parallellabs.app/api-reference/browser/get-browser-task-status-and-result

/openapi.json get /api/v0/browser-task/{task_id}
Poll this endpoint to check the status of a submitted browser task.
When status is 'completed', the result will contain the task output.
When status is 'failed', the errorMessage will contain the failure reason.



# List available browser integrations.
Source: https://docs.parallellabs.app/api-reference/browser/list-available-browser-integrations

/openapi.json get /api/v0/browser-integrations
Returns a list of browser session integrations (LinkedIn, Twitter, etc.)
that can be used for authenticated browser tasks.



# Submit a browser automation task.
Source: https://docs.parallellabs.app/api-reference/browser/submit-a-browser-automation-task

/openapi.json post /api/v0/browser-task
Submits a browser task for asynchronous execution. Returns a task ID
that can be used to poll for status and results.

Supports three session types:
- regular: Standard browser session
- authenticated: Uses an integration for logged-in sessions
- residential: Uses a residential proxy with specified zipcode



# Generate custom music.
Source: https://docs.parallellabs.app/api-reference/canvas/generate-custom-music

/openapi.json post /api/v0/audio/music
Default to free stock music for music, soundtracks, reels and stories: use search_stock_music then import_stock_music. Use paid create_music ONLY when the user explicitly requests custom AI-generated music. Never automatically fall back to paid generation when stock search/import fails; try different keywords or tracks and report if none are suitable.
Create music with ElevenLabs and return a permanent public audio URL and duration in seconds. Describe genre, mood, instruments, tempo and any desired lyrics. Instrumental defaults to true for background music underneath narration. Embed the URL in animated canvas HTML as <audio src="URL" loop data-volume="0.2"></audio> to include it in the exported MP4. Generation consumes audio credits.



# Import licensed stock music into the audio library.
Source: https://docs.parallellabs.app/api-reference/canvas/import-licensed-stock-music-into-the-audio-library

/openapi.json post /api/v0/stock-music/import
Import a selected search_stock_music result by stockAudioId. Rechecks its indexed license and matching license link on the original source page, saves an MP3 with source/license metadata, and returns a permanent public audioUrl plus measured duration in seconds. Use audioUrl in <audio src="URL" loop data-volume="0.2"></audio> in an animated canvas. No audio generation credits are charged. Sources that cannot be verified are rejected; choose another track. Imports are bounded by the configured file-size and duration limits (default 40 MiB and the first ten minutes).



# Render a canvas art image or animation from HTML
Source: https://docs.parallellabs.app/api-reference/canvas/render-a-canvas-art-image-or-animation-from-html

/openapi.json post /api/v0/canvas
Renders a self-contained HTML canvas at exact pixel dimensions and returns a permanent public URL — for social posts, story graphics, quote cards, and link preview art. Costs 0.1 credits per call.

Pass animated=true to record the canvas as a looping MP4 instead of a still PNG (animate with infinite CSS keyframes); that returns videoId and an MP4 url, and costs 3 credits.

Author the HTML to these design guidelines:

## What You Are Making
A canvas is a piece of visual art, not a document. Museum-quality, magazine-quality —
never cartoony, never clip-art, never a slide. Before writing any HTML, decide a design
direction in one sentence (e.g. "Swiss formalism: one monumental word, hard grid,
red/black/paper" or "quiet editorial: thin serif, vast negative space, one photograph").
Every choice then serves that direction.

## Brand Style (check first)
- Before choosing a direction, check whether the company has a brand kit saved in
  the content engine (GET /brand-kits — brand_kit_listBrandKits on the platform
  API/MCP). Do this BEFORE the first render, not after.
- If a brand kit exists, its `imageStylePrompt` is the company's established visual
  style (palette, mood, typography feel, photographic treatment); its `images` list
  holds brand assets (logos, product shots, approved photography) with public URLs;
  and its `webArtifacts` holds the primary `favicon` and `logo` (each with a public
  `url`). Treat these as the source of truth: derive the palette and type choices
  from the style prompt, and embed brand assets by their exact URL instead of stock
  or AI-generated stand-ins when a suitable one exists. Never alter, recolor, or
  distort a logo.
- If no brand kit exists, or the session already supplies brand context in its
  instructions, skip the lookup and design from the brief.

## Canvas Setup (exact rules)
- Output one complete standalone HTML document. No external CSS/JS files; inline a
  `<style>` block. Plain CSS preferred, but these shared platform libraries may be loaded
  from their CDNs:
  - Tailwind CSS: `<script src="https://cdn.tailwindcss.com"></script>` — utility classes; no build step
  - Lucide Icons: `<script src="https://unpkg.com/lucide@1.33.0/dist/umd/lucide.min.js"></script>` — place `<i data-lucide="arrow-right"></i>`, then call `lucide.createIcons();` after the last icon
  - Google Fonts: `<link href="https://fonts.googleapis.com/css2?family=...&display=swap" rel="stylesheet">` — pick faces that match the brand
- Size the canvas EXACTLY to the chosen format:
  `html,body{margin:0;padding:0}` and `body{width:<W>px;height:<H>px;overflow:hidden;position:relative}`
- Named formats include: square 1080x1080 (feed), portrait 1080x1350 (feed, default), story
  1080x1920 (Stories/Reels), landscape 1200x630 (link previews), slide 1920x1080 (16:9
  presentation slide background — decorative art only, no text; editable text is layered on
  later), plus profile covers (linkedin_banner, twitter_header, facebook_cover),
  youtube_thumb, email_header, print pages (a4_portrait, a4_landscape, letter_portrait) and
  business_card. When none fits, an exact custom width/height (200-2160px each) can be
  supplied instead — size the body to whatever exact dimensions the session gives you.
- A canvas is rendered one of two ways, and you are told which one this session wants:
  - STILL (default): the canvas is photographed the instant it loads. Design it static —
    no animations or transitions, since only the first frame is ever seen.
  - ANIMATED: the canvas is recorded to a short MP4. Follow the animation rules below.

## Typography (the heart of the canvas)
- Google Fonts only, loaded via `<link href="https://fonts.googleapis.com/css2?family=...&display=swap" rel="stylesheet">`.
- Choose from this curated library — at most TWO fonts per canvas:
  - Monumental display: Boldonse, Big Shoulders, Erica One, Archivo Black, Bebas Neue
  - Elegant serif: Gloock, Italiana, Young Serif, Instrument Serif, Playfair Display, Libre Baskerville
  - Modern sans: Bricolage Grotesque, Outfit, Work Sans, Instrument Sans, Space Grotesk
  - Editorial body serif: Lora, Crimson Pro, IBM Plex Serif
  - Technical mono: JetBrains Mono, IBM Plex Mono, DM Mono, Tektur, Jura
  - Retro pixel: Silkscreen, Pixelify Sans
  - Handwritten: Nothing You Could Do, Smooch Sans; thin deco: Poiret One
- Text is minimal and visual-first: usually one dominant word or short phrase plus tiny
  supporting labels. Rarely more than a dozen words on the whole canvas.
- Make the hero text HUGE — roughly 10-25% of the canvas height. Uppercase display type
  wants letter-spacing (0.02-0.15em); tiny labels want wide tracking (0.2-0.4em).
- Line-height on stacked display lines: some display fonts draw far taller than their
  em box and collide when stacked (Boldonse needs line-height ~1.3). Use 1.2-1.3 for
  multi-line display type and only go tighter for fonts you know sit inside their box.
- Contrast in scale is the composition: one enormous element, a few whisper-small ones.
  Nothing in between.

## Icons (Lucide)
- When the design calls for an icon — a directional arrow, a play glyph, a small
  functional mark next to a label — use Lucide, never emoji and never clip-art.
- Load it once, at the END of the body, then render the icons:
  `<script src="https://unpkg.com/lucide@1.33.0/dist/umd/lucide.min.js"></script>` followed by `<script>lucide.createIcons();</script>`
- Place an icon with `<i data-lucide="arrow-right"></i>`. Size and color it through CSS on
  the element — Lucide SVGs inherit `currentColor` and take `width`/`height`:
  `i[data-lucide]{width:64px;height:64px;stroke-width:1.5}` (the stroke weight matters as
  much as the size — hairline strokes read as refined, thick ones as cartoonish).
- Use real Lucide names (arrow-right, arrow-up-right, sparkles, zap, check, star, play,
  quote, trending-up, circle-dot). A misspelled name renders nothing at all.
- Icons are accents, not the subject: one or two per canvas, small against the hero type.
  An icon never substitutes for a real design idea.

## Animation (ANIMATED canvases only)
Only when the session asks for an animated canvas. A still canvas must have no animation
at all — it is photographed on the first frame.
- Animate with CSS `@keyframes` on the elements themselves. The recording seeks the page's
  animation timeline frame by frame, so what is recorded is exactly your CSS timeline.
- Every animation MUST be `infinite` and MUST tile seamlessly: the state at the end of one
  iteration has to match the state at the start, or the loop visibly jumps. The cleanest
  way is a single cycle that returns home (`0%` and `100%` identical).
- Give the whole piece ONE shared cycle length — set every animation to the same
  `animation-duration` (or an exact divisor of it, e.g. 2s inside an 8s piece) so the
  composition repeats as a whole rather than drifting out of phase.
- Stagger with `animation-delay` (negative delays start an element mid-cycle), never by
  giving elements unrelated durations.
- Animate `transform` and `opacity` above all — they are what animates smoothly. Also fine:
  `filter`, `clip-path`, gradient positions, `stroke-dashoffset` on SVG.
- Keep it restrained and slow: a drifting gradient, a slow parallax push, type that breathes,
  a single element sliding through. Elegant motion is subtle — nothing bouncing, spinning
  fast, or flashing. The canvas must still look composed in any single frozen frame.

### What the recorder can and cannot capture
The recording pauses the browser's animation timeline and steps through it — it does NOT
watch the page play in real time. Only animations the browser itself drives are captured:
- CAPTURED: CSS `@keyframes`, CSS transitions, SVG SMIL, and the Web Animations API
  (`element.animate(...)`) — everything that appears in `document.getAnimations()`.
  HTML `<video>` frames are also seeked in step with the animation timeline.
- NOT CAPTURED (records as a frozen video, with no error): anything driven by JavaScript
  frame by frame — `requestAnimationFrame`, `setInterval`/`setTimeout` loops, canvas 2D/WebGL
  draw loops, and scroll-triggered libraries. Animated GIFs are not seeked reliably.
- This is the ONE rule that differs from the landing page builder, whose pages play live in
  a real browser and can use anything. Do not carry those habits over.

### Stock video for animated canvases — search_stock_videos
Search Pexels stock videos for animated canvases. Returns direct videoUrl MP4 URLs, preview images, dimensions, duration and creator attribution. Use the exact videoUrl in <video autoplay loop muted playsinline> with object-fit:cover, sized to the canvas; layer readable text above it. Set animated=true when calling create_canvas, or set_animated(True) in the canvas builder. Video frames are seeked with the animation timeline in exports; use separate audio tracks for sound. Still PNGs have no motion. Paginate with page/pageSize; this tool is available whenever canvas tools are enabled.

### Audio for animated canvases
Default to free stock music for music, soundtracks, reels and stories: use search_stock_music then import_stock_music. Use paid create_music ONLY when the user explicitly requests custom AI-generated music. Never automatically fall back to paid generation when stock search/import fails; try different keywords or tracks and report if none are suitable.
Search Openverse for free CC0/public-domain stock music and loops by mood, genre or instruments. Returns candidates with stockAudioId, previewUrl, duration in seconds, creator and source/license links. Choose a track and call import_stock_music to get a permanent canvas audio URL. Do not embed previewUrl in a canvas. Search consumes no audio credits.
Import a selected search_stock_music result by stockAudioId. Rechecks its indexed license and matching license link on the original source page, saves an MP3 with source/license metadata, and returns a permanent public audioUrl plus measured duration in seconds. Use audioUrl in <audio src="URL" loop data-volume="0.2"></audio> in an animated canvas. No audio generation credits are charged. Sources that cannot be verified are rejected; choose another track. Imports are bounded by the configured file-size and duration limits (default 40 MiB and the first ten minutes).

- For narration or dialogue, call `create_speech`. Audio tools return permanent public
  audio URLs and measured durations. Never invent audio URLs. If speech generation
  is unavailable, explain that limitation.
- Include tracks directly in the HTML: `<audio src="RETURNED_URL" loop
  data-volume="0.2"></audio>` for quiet background music; `<audio src="RETURNED_URL"
  data-start="2" data-volume="1"></audio>` for narration starting 2 seconds in.
- Up to 8 tracks are mixed into the exported MP4. data-start is seconds (default 0);
  data-volume is 0..1 (default 1). `loop` repeats music; omit it for speech. `muted`
  excludes a track. Only platform public generated or imported audio URLs are accepted.
- For dialogue, generate each speaker's line separately with the desired voice ID,
  then use data-start and the returned durations to place the lines in sequence.
  Lower the music volume under dialogue. Do not repeat or cut off spoken lines.
- Set video duration long enough for all lines, within the tool's duration limit;
  shorten the script if needed. With audio the requested duration is preserved,
  without rounding to whole animation cycles. Tracks are trimmed/padded to that length.
- Audio is mixed during export; browser autoplay, Web Audio and JavaScript playback
  are not recorded. HTML preview autoplay may be blocked; use the exported MP4 to
  review the finished soundtrack. Still PNGs have no audio.

### Animation libraries
CSS keyframes are the default and are always enough for a canvas — reach for a library only
when the motion genuinely needs orchestration (staggered sequences, timelines).
- The ONLY animation library available here is Anime.js, and ONLY through its WAAPI entry
  point, which puts animations on the browser timeline:
  - Anime.js (advanced animation): `<script src="https://cdn.jsdelivr.net/npm/animejs/dist/bundles/anime.umd.min.js"></script>` — staggered reveals, timelines, kinetic type. `anime.waapi.animate(...)` puts the animation on the browser timeline; `anime.animate(...)` / `anime.Timeline` drive it from JS frame by frame
  Call it as `anime.waapi.animate('.el', { translateX: 300, duration: 2000, loop: true })`.
  Its classic `anime.animate(...)` / `anime.Timeline` API is JS-driven and records as a
  FROZEN video — if you use Anime.js at all, every animation must go through
  `anime.waapi.animate`.
- AOS (used by the landing page builder) is scroll-triggered and USELESS here: a canvas is a
  fixed viewport with nothing to scroll, so its animations never fire. Never load it.
- Whatever you use, the seamless-loop and shared-cycle rules above still apply.

## Color & Background
- A limited, intentional palette: 2-3 colors total. Pick them for the design direction,
  not decoration. High text/background contrast is non-negotiable.
- Backgrounds: a flat color, a subtle gradient, or a photo. For photos use
  `background:url(...) center/cover` and ALWAYS add a scrim over photo areas behind text
  (e.g. `background:rgba(0,0,0,.35)` or a directional gradient) so type stays legible.
  The scrim serves the text, not the whole canvas: keep it light (≤ ~0.5 opacity) or
  directional so the photo stays clearly visible — a photo buried under a near-opaque
  overlay reads as a plain dark field and wastes the imagery.
- Photo URLs must be public http(s) images. Two ways to get one:
  - Stock photography: search the stock image library (Unsplash/Pexels via the stock
    image search — images_searchStockImages on the platform API/MCP) and use a
    result's fullUrl.
  - AI-generated imagery: generate an image with the platform's AI image models
    (images_createImage on the platform API/MCP — describe the scene, pick a model,
    and use the returned public URL). Best for imagery no stock photo can match;
    request it WITHOUT text, since the canvas type layer supplies all text.

## Composition (non-negotiable craft rules)
- Keep a safe margin of at least 5% of the canvas on every side. NOTHING touches or
  overflows the canvas edge, and elements never overlap illegibly.
- Compose deliberately: strong asymmetry or a strict grid both work; accidental
  centering of everything does not. Let negative space breathe — empty space is a
  design element, not waste.
- Repetition rewards attention: thin rules, small registration marks, an index number,
  a tiny caption line — sparse, clinical details make the piece feel designed.
- The finished canvas should look like it took a top-of-field designer hours of
  painstaking refinement: aligned edges, consistent spacing rhythm, optically balanced.
  Double-check every element is inside the canvas and nothing collides before finishing.




# Search free stock music.
Source: https://docs.parallellabs.app/api-reference/canvas/search-free-stock-music

/openapi.json get /api/v0/stock-music/search
Default to free stock music for music, soundtracks, reels and stories: use search_stock_music then import_stock_music. Use paid create_music ONLY when the user explicitly requests custom AI-generated music. Never automatically fall back to paid generation when stock search/import fails; try different keywords or tracks and report if none are suitable.
Search Openverse for free CC0/public-domain stock music and loops by mood, genre or instruments. Returns candidates with stockAudioId, previewUrl, duration in seconds, creator and source/license links. Choose a track and call import_stock_music to get a permanent canvas audio URL. Do not embed previewUrl in a canvas. Search consumes no audio credits.



# Search stock videos.
Source: https://docs.parallellabs.app/api-reference/canvas/search-stock-videos

/openapi.json get /api/v0/stock-videos/search
Search Pexels stock videos for animated canvases. Returns direct videoUrl MP4 URLs, preview images, dimensions, duration and creator attribution. Use the exact videoUrl in <video autoplay loop muted playsinline> with object-fit:cover, sized to the canvas; layer readable text above it. Set animated=true when calling create_canvas, or set_animated(True) in the canvas builder. Video frames are seeked with the animation timeline in exports; use separate audio tracks for sound. Still PNGs have no motion. Paginate with page/pageSize; this tool is available whenever canvas tools are enabled.



# Count tokens for a message request (Anthropic API compatible).
Source: https://docs.parallellabs.app/api-reference/chat/count-tokens-for-a-message-request-anthropic-api-compatible

/openapi.json post /api/v0/claude/v1/messages/count_tokens
Character-based estimate; close enough for Claude Code's context
management.



# Create a chat completion (OpenAI compatible).
Source: https://docs.parallellabs.app/api-reference/chat/create-a-chat-completion-openai-compatible

/openapi.json post /api/v0/chat/completions
Fully compatible with OpenAI's API specification for seamless
integration with OpenAI libraries. Supports both streaming and
non-streaming responses.



# Create a message (Anthropic Messages API compatible).
Source: https://docs.parallellabs.app/api-reference/chat/create-a-message-anthropic-messages-api-compatible

/openapi.json post /api/v0/claude/v1/messages
Point Claude Code or any Anthropic SDK at
ANTHROPIC_BASE_URL=https://<your-domain>/api/v0/claude with a platform
API key as ANTHROPIC_AUTH_TOKEN. Supports streaming and tool use.



# List available models (Anthropic API compatible).
Source: https://docs.parallellabs.app/api-reference/chat/list-available-models-anthropic-api-compatible

/openapi.json get /api/v0/claude/v1/models
Lets Claude Code discover platform models when
CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERY=1 is set. This API serves every
request with ANTHROPIC_COMPAT_MODEL, so that is the only model listed.



# Create or update a company.
Source: https://docs.parallellabs.app/api-reference/companies/create-or-update-a-company

/openapi.json post /api/v0/companies



# Delete a company.
Source: https://docs.parallellabs.app/api-reference/companies/delete-a-company

/openapi.json delete /api/v0/companies/{companyId}



# Get a specific company by ID.
Source: https://docs.parallellabs.app/api-reference/companies/get-a-specific-company-by-id

/openapi.json get /api/v0/companies/{companyId}



# Get all companies accessible by the authenticated user.
Source: https://docs.parallellabs.app/api-reference/companies/get-all-companies-accessible-by-the-authenticated-user

/openapi.json get /api/v0/companies



# Get all tags for a company.
Source: https://docs.parallellabs.app/api-reference/companies/get-all-tags-for-a-company

/openapi.json get /api/v0/companies/{companyId}/tags



# Remove the current user from a company. Billing owner cannot leave.
Source: https://docs.parallellabs.app/api-reference/companies/remove-the-current-user-from-a-company-billing-owner-cannot-leave

/openapi.json post /api/v0/companies/{companyId}/leave



# Transfer company ownership to another company owner.
Source: https://docs.parallellabs.app/api-reference/companies/transfer-company-ownership-to-another-company-owner

/openapi.json put /api/v0/companies/{companyId}/transfer-ownership



# Update allowed features for a company (owner only).
Source: https://docs.parallellabs.app/api-reference/companies/update-allowed-features-for-a-company-owner-only

/openapi.json put /api/v0/companies/{companyId}/features



# Update the display order of companies.
Source: https://docs.parallellabs.app/api-reference/companies/update-the-display-order-of-companies

/openapi.json post /api/v0/companies/order



# Add a user/collaborator to a company.
Source: https://docs.parallellabs.app/api-reference/company-users/add-a-usercollaborator-to-a-company

/openapi.json post /api/v0/company_users



# Get all users/collaborators for a company with pagination.
Source: https://docs.parallellabs.app/api-reference/company-users/get-all-userscollaborators-for-a-company-with-pagination

/openapi.json get /api/v0/company_users/{companyId}



# Get the current user's role in a specific company.
Source: https://docs.parallellabs.app/api-reference/company-users/get-the-current-users-role-in-a-specific-company

/openapi.json get /api/v0/company_users/{companyId}/my-role



# Get the subscription type of a company member by their ID.
Source: https://docs.parallellabs.app/api-reference/company-users/get-the-subscription-type-of-a-company-member-by-their-id

/openapi.json get /api/v0/company_users/{memberId}/subscription-type



# Remove a user/collaborator from a company.
Source: https://docs.parallellabs.app/api-reference/company-users/remove-a-usercollaborator-from-a-company

/openapi.json delete /api/v0/company_users/{memberId}



# Resend an invitation email to a collaborator who hasn't logged in yet.
Source: https://docs.parallellabs.app/api-reference/company-users/resend-an-invitation-email-to-a-collaborator-who-hasnt-logged-in-yet

/openapi.json post /api/v0/company_users/{memberId}/resend-invite



# Update a company user's role.
Source: https://docs.parallellabs.app/api-reference/company-users/update-a-company-users-role

/openapi.json put /api/v0/company_users/{memberId}



# Update the allowed features for a company user.
Source: https://docs.parallellabs.app/api-reference/company-users/update-the-allowed-features-for-a-company-user

/openapi.json put /api/v0/company_users/{memberId}/features



# Create a content research project.
Source: https://docs.parallellabs.app/api-reference/content-research/create-a-content-research-project

/openapi.json post /api/v0/research/projects
Create one project per ideal customer profile or product line. Each run
analyses that project's domain, competitors and ICP on its own, so
separate audiences stay separated rather than being blended into one
result set.



# Delete a research project.
Source: https://docs.parallellabs.app/api-reference/content-research/delete-a-research-project

/openapi.json delete /api/v0/research/projects/{projectId}
Soft-deletes the project. Its opportunities stop appearing in listings.



# Dismiss a content opportunity.
Source: https://docs.parallellabs.app/api-reference/content-research/dismiss-a-content-opportunity

/openapi.json post /api/v0/research/opportunities/{opportunityId}/dismiss
Hides it from the default listing. Re-running the project will not
resurface it as new.



# Generate a content piece from an opportunity's brief.
Source: https://docs.parallellabs.app/api-reference/content-research/generate-a-content-piece-from-an-opportunitys-brief

/openapi.json post /api/v0/research/opportunities/{opportunityId}/generate
Creates a record in the Content Engine and writes the piece in the
background using the opportunity's title, outline, target keywords and
AEO recommendations. Defaults to the parent project's ICP unless icpId
is supplied. Returns immediately; the opportunity moves to "generated".



# Get a single content opportunity.
Source: https://docs.parallellabs.app/api-reference/content-research/get-a-single-content-opportunity

/openapi.json get /api/v0/research/opportunities/{opportunityId}
Returns the full record including the score breakdown, generation brief
and supporting SERP and AI-citation evidence.



# List all content research projects for a company.
Source: https://docs.parallellabs.app/api-reference/content-research/list-all-content-research-projects-for-a-company

/openapi.json get /api/v0/research/projects
Each project targets one domain and one ideal customer profile. A company
can have many, and opportunities are scoped per project.



# List run history for a research project.
Source: https://docs.parallellabs.app/api-reference/content-research/list-run-history-for-a-research-project

/openapi.json get /api/v0/research/projects/{projectId}/runs
Newest first. Poll this after triggering a run: the first entry carries
the live status and phase while the pipeline is working.



# List scored content opportunities.
Source: https://docs.parallellabs.app/api-reference/content-research/list-scored-content-opportunities

/openapi.json get /api/v0/research/opportunities
Pass projectId to scope results to one project's ICP; omit it to see
opportunities across every project for the company.



# Restore a dismissed opportunity.
Source: https://docs.parallellabs.app/api-reference/content-research/restore-a-dismissed-opportunity

/openapi.json post /api/v0/research/opportunities/{opportunityId}/restore
Returns it to "new" so it appears in the default listing again.



# Start a research run for a project.
Source: https://docs.parallellabs.app/api-reference/content-research/start-a-research-run-for-a-project

/openapi.json post /api/v0/research/projects/{projectId}/run
Runs asynchronously through the SERP, AI-answer, keyword-metrics,
domain-authority and analysis phases, then upserts scored opportunities.
Poll the runs endpoint for status. Only one run per project can be active
at a time - triggering while one is in flight returns that run with
started=false rather than starting a second.

Charges a flat credit cost up front, which is not refunded if the run
later fails.



# Suggest setup values for a new research project.
Source: https://docs.parallellabs.app/api-reference/content-research/suggest-setup-values-for-a-new-research-project

/openapi.json get /api/v0/research/prefill
Infers a project name, competitor domains and seed keywords from the
company's own website and profile, so a project can be created without
the caller supplying them.



# Track an opportunity's keywords with a monitor.
Source: https://docs.parallellabs.app/api-reference/content-research/track-an-opportunitys-keywords-with-a-monitor

/openapi.json post /api/v0/research/opportunities/{opportunityId}/track
Creates a daily keyword monitor across Google AI and ChatGPT for up to
the opportunity's first three target keywords, so citation changes are
picked up over time. Fails if the opportunity is already tracked.



# Update a research project's settings.
Source: https://docs.parallellabs.app/api-reference/content-research/update-a-research-projects-settings

/openapi.json patch /api/v0/research/projects/{projectId}
Changing the domain, competitors, seed keywords or ICP affects the next
run; it does not re-score opportunities already found.



# Analyze email content for spam triggers and deliverability issues.
Source: https://docs.parallellabs.app/api-reference/content/analyze-email-content-for-spam-triggers-and-deliverability-issues

/openapi.json post /api/v0/content/spam-check
Based on email deliverability best practices and spam filter guidelines,
this endpoint analyzes email subject and content for potential issues.



# Ask the thread's in-flight agent run to wrap up (same signal a new
Source: https://docs.parallellabs.app/api-reference/content/ask-the-threads-in-flight-agent-run-to-wrap-up-same-signal-a-new

/openapi.json post /api/v0/content/canvases/{canvas_id}/threads/{thread_id}/stop
message sends; the partial reply is already saved).



# Assign content to a collaborator for review/edit/approval.
Source: https://docs.parallellabs.app/api-reference/content/assign-content-to-a-collaborator-for-revieweditapproval

/openapi.json post /api/v0/content/{contentId}/assign



# Bulk schedule multiple content items with the same publish account.
Source: https://docs.parallellabs.app/api-reference/content/bulk-schedule-multiple-content-items-with-the-same-publish-account

/openapi.json post /api/v0/content/bulk-schedule



# Chat endpoint for the AI canvas builder.
Source: https://docs.parallellabs.app/api-reference/content/chat-endpoint-for-the-ai-canvas-builder

/openapi.json post /api/v0/content/canvases/{canvas_id}/chat
Shares the app builder's agent loop and typed SSE protocol (text /
tool_call / tool_result / files_changed / done / error). The thread is the
source of truth: history is loaded from the canvas's stored threads, the
user message is persisted immediately, and the assistant message (with its
tool-call log) is persisted when the stream ends.



# Clone a public prompt to the user's company.
Source: https://docs.parallellabs.app/api-reference/content/clone-a-public-prompt-to-the-users-company

/openapi.json post /api/v0/content/prompts/{promptId}/clone



# Convert text content to spintax format using AI.
Source: https://docs.parallellabs.app/api-reference/content/convert-text-content-to-spintax-format-using-ai

/openapi.json post /api/v0/content/spintax
Spintax allows for variation in messages using {{RANDOM | option1 | option2}} syntax.



# Create a new canvas design.
Source: https://docs.parallellabs.app/api-reference/content/create-a-new-canvas-design

/openapi.json post /api/v0/content/canvases



# Create a new chat thread (becomes the active thread).
Source: https://docs.parallellabs.app/api-reference/content/create-a-new-chat-thread-becomes-the-active-thread

/openapi.json post /api/v0/content/canvases/{canvas_id}/threads



# Create or update a brand voice.
Source: https://docs.parallellabs.app/api-reference/content/create-or-update-a-brand-voice

/openapi.json post /api/v0/content/brand-voices



# Create or update a content campaign.
Source: https://docs.parallellabs.app/api-reference/content/create-or-update-a-content-campaign

/openapi.json post /api/v0/content/campaigns



# Create or update a content strategy.
Source: https://docs.parallellabs.app/api-reference/content/create-or-update-a-content-strategy

/openapi.json post /api/v0/content/strategy



# Create or update a prompt.
Source: https://docs.parallellabs.app/api-reference/content/create-or-update-a-prompt

/openapi.json post /api/v0/content/prompts



# Create or update an ICP profile.
Source: https://docs.parallellabs.app/api-reference/content/create-or-update-an-icp-profile

/openapi.json post /api/v0/content/profiles



# Delete a brand voice.
Source: https://docs.parallellabs.app/api-reference/content/delete-a-brand-voice

/openapi.json delete /api/v0/content/brand-voices/{voiceId}



# Delete a canvas design.
Source: https://docs.parallellabs.app/api-reference/content/delete-a-canvas-design

/openapi.json delete /api/v0/content/canvases/{canvas_id}



# Delete a content campaign.
Source: https://docs.parallellabs.app/api-reference/content/delete-a-content-campaign

/openapi.json delete /api/v0/content/campaigns/{contentCampaignId}



# Delete a content item.
Source: https://docs.parallellabs.app/api-reference/content/delete-a-content-item

/openapi.json delete /api/v0/content/{contentId}



# Delete a content strategy.
Source: https://docs.parallellabs.app/api-reference/content/delete-a-content-strategy

/openapi.json delete /api/v0/content/strategy/{strategyId}



# Delete a prompt.
Source: https://docs.parallellabs.app/api-reference/content/delete-a-prompt

/openapi.json delete /api/v0/content/prompts/{promptId}



# Delete a thread; returns the refreshed thread view.
Source: https://docs.parallellabs.app/api-reference/content/delete-a-thread;-returns-the-refreshed-thread-view

/openapi.json delete /api/v0/content/canvases/{canvas_id}/threads/{thread_id}



# Delete an ICP profile.
Source: https://docs.parallellabs.app/api-reference/content/delete-an-icp-profile

/openapi.json delete /api/v0/content/profiles/{profileId}



# Download content as PDF.
Source: https://docs.parallellabs.app/api-reference/content/download-content-as-pdf

/openapi.json get /api/v0/content/{contentId}/pdf



# Export content ideas to CSV format.
Source: https://docs.parallellabs.app/api-reference/content/export-content-ideas-to-csv-format

/openapi.json post /api/v0/content/export



# Generate a brand voice from a LinkedIn profile (person or company).
Source: https://docs.parallellabs.app/api-reference/content/generate-a-brand-voice-from-a-linkedin-profile-person-or-company

/openapi.json post /api/v0/content/brand-voices/generate-from-linkedin
Either linkedin_url or link_id must be provided. The profile_type will be
auto-detected from the URL if not specified.



# Generate a brand voice using AI based on input data.
Source: https://docs.parallellabs.app/api-reference/content/generate-a-brand-voice-using-ai-based-on-input-data

/openapi.json post /api/v0/content/brand-voices/generate



# Generate a content strategy using AI based on input data.
Source: https://docs.parallellabs.app/api-reference/content/generate-a-content-strategy-using-ai-based-on-input-data

/openapi.json post /api/v0/content/strategy/generate



# Generate a short AI snippet based on a prompt with variable placeholders.
Source: https://docs.parallellabs.app/api-reference/content/generate-a-short-ai-snippet-based-on-a-prompt-with-variable-placeholders

/openapi.json post /api/v0/content/ai-snippet
This endpoint is used at send time to generate small, personalized text snippets.
The AI has guardrails to keep responses short (1-2 sentences max).



# Generate and attach agent-designed canvases using a selected brand kit.
Source: https://docs.parallellabs.app/api-reference/content/generate-and-attach-agent-designed-canvases-using-a-selected-brand-kit

/openapi.json post /api/v0/content/{contentId}/canvas



# Generate content from a content idea.
Source: https://docs.parallellabs.app/api-reference/content/generate-content-from-a-content-idea

/openapi.json post /api/v0/content/generate-from-idea



# Generate content using AI based on context and user instructions.
Source: https://docs.parallellabs.app/api-reference/content/generate-content-using-ai-based-on-context-and-user-instructions

/openapi.json post /api/v0/content/generate
Supports email, sms, linkedin_message, and voice_message content types.



# Get a canvas design (includes the current HTML for the preview).
Source: https://docs.parallellabs.app/api-reference/content/get-a-canvas-design-includes-the-current-html-for-the-preview

/openapi.json get /api/v0/content/canvases/{canvas_id}



# Get a specific brand voice by ID.
Source: https://docs.parallellabs.app/api-reference/content/get-a-specific-brand-voice-by-id

/openapi.json get /api/v0/content/brand-voices/{voiceId}



# Get a specific content strategy by ID.
Source: https://docs.parallellabs.app/api-reference/content/get-a-specific-content-strategy-by-id

/openapi.json get /api/v0/content/strategy/{strategyId}



# Get a specific ICP profile by ID.
Source: https://docs.parallellabs.app/api-reference/content/get-a-specific-icp-profile-by-id

/openapi.json get /api/v0/content/profiles/{profileId}



# Get a specific prompt by ID.
Source: https://docs.parallellabs.app/api-reference/content/get-a-specific-prompt-by-id

/openapi.json get /api/v0/content/prompts/{promptId}



# Get all images associated with a specific content.
Source: https://docs.parallellabs.app/api-reference/content/get-all-images-associated-with-a-specific-content

/openapi.json get /api/v0/content/{contentId}/images



# Get content assigned to the current user or all assigned content.
Source: https://docs.parallellabs.app/api-reference/content/get-content-assigned-to-the-current-user-or-all-assigned-content

/openapi.json get /api/v0/content/assigned



# Get one thread's messages.
Source: https://docs.parallellabs.app/api-reference/content/get-one-threads-messages

/openapi.json get /api/v0/content/canvases/{canvas_id}/threads/{thread_id}



# Get overall statistics for content engine.
Source: https://docs.parallellabs.app/api-reference/content/get-overall-statistics-for-content-engine

/openapi.json get /api/v0/content/statistics



# Get the default content strategy for a company.
Source: https://docs.parallellabs.app/api-reference/content/get-the-default-content-strategy-for-a-company

/openapi.json get /api/v0/content/strategy



# Humanize AI-generated text so it reads naturally human-written.
Source: https://docs.parallellabs.app/api-reference/content/humanize-ai-generated-text-so-it-reads-naturally-human-written

/openapi.json post /api/v0/content/humanize
Strips invisible/homoglyph characters, then runs the text through the
content engine's AI-tell checks (AI-typical phrases, uniform sentence
length, em dashes, readability) and rewrites it against the returned
guidance, preserving meaning, facts, formatting, and approximate length.
Optional instructions (e.g. "keep it casual") take precedence over the
default rules. Billed against the content-engine model quota.



# Import content ideas from CSV data.
Source: https://docs.parallellabs.app/api-reference/content/import-content-ideas-from-csv-data

/openapi.json post /api/v0/content/import



# List all brand voices for a company.
Source: https://docs.parallellabs.app/api-reference/content/list-all-brand-voices-for-a-company

/openapi.json get /api/v0/content/brand-voices



# List all content campaigns.
Source: https://docs.parallellabs.app/api-reference/content/list-all-content-campaigns

/openapi.json get /api/v0/content/campaigns



# List all content items with optional filtering.
Source: https://docs.parallellabs.app/api-reference/content/list-all-content-items-with-optional-filtering

/openapi.json get /api/v0/content



# List all content strategies for a company.
Source: https://docs.parallellabs.app/api-reference/content/list-all-content-strategies-for-a-company

/openapi.json get /api/v0/content/strategies



# List all ICP profiles for a company.
Source: https://docs.parallellabs.app/api-reference/content/list-all-icp-profiles-for-a-company

/openapi.json get /api/v0/content/profiles



# List all public prompts from any company.
Source: https://docs.parallellabs.app/api-reference/content/list-all-public-prompts-from-any-company

/openapi.json get /api/v0/content/prompts/public



# List canvas designs for a company (paginated).
Source: https://docs.parallellabs.app/api-reference/content/list-canvas-designs-for-a-company-paginated

/openapi.json get /api/v0/content/canvases



# List chat threads for a canvas ({threads, activeThreadId}).
Source: https://docs.parallellabs.app/api-reference/content/list-chat-threads-for-a-canvas-

/openapi.json get /api/v0/content/canvases/{canvas_id}/threads



# List prompts visible to the user (own private + shared in company).
Source: https://docs.parallellabs.app/api-reference/content/list-prompts-visible-to-the-user-own-private-+-shared-in-company

/openapi.json get /api/v0/content/prompts



# Manually create a new content engine record.
Source: https://docs.parallellabs.app/api-reference/content/manually-create-a-new-content-engine-record

/openapi.json post /api/v0/content
Use this to insert a content piece directly (skipping AI ideation/generation),
e.g. to import an existing post or log content authored elsewhere. The body
is set via either `variants.default` or the `body` convenience field.



# Publish content to a connected account.
Source: https://docs.parallellabs.app/api-reference/content/publish-content-to-a-connected-account

/openapi.json post /api/v0/content/{contentId}/publish



# Record the canvas's saved HTML to a permanent public looping MP4 (charged at DATA_CREDIT_COST_CANVAS_VIDEO).
Source: https://docs.parallellabs.app/api-reference/content/record-the-canvass-saved-html-to-a-permanent-public-looping-mp4-charged-at-data_credit_cost_canvas_video

/openapi.json post /api/v0/content/canvases/{canvas_id}/render-video
Seeks the design's CSS animation timeline frame by frame, so the canvas
HTML must animate with CSS keyframes for the video to show any motion.



# Rename a thread (`title`) or make it active (`active: true`).
Source: https://docs.parallellabs.app/api-reference/content/rename-a-thread-`title`-or-make-it-active-`active:-true`

/openapi.json patch /api/v0/content/canvases/{canvas_id}/threads/{thread_id}



# Render the canvas's saved HTML to a permanent public PNG (charged at DATA_CREDIT_COST_CANVAS).
Source: https://docs.parallellabs.app/api-reference/content/render-the-canvass-saved-html-to-a-permanent-public-png-charged-at-data_credit_cost_canvas

/openapi.json post /api/v0/content/canvases/{canvas_id}/render



# Save/update a content item.
Source: https://docs.parallellabs.app/api-reference/content/saveupdate-a-content-item

/openapi.json post /api/v0/content/{contentId}



# Score content against its focus keyword with the formula-based SEO checks.
Source: https://docs.parallellabs.app/api-reference/content/score-content-against-its-focus-keyword-with-the-formula-based-seo-checks

/openapi.json post /api/v0/content/seo-score
Pure text analysis (the same checks WordPress SEO plugins run) - no model
call, so it is safe to invoke on every edit. Returns the overall 0-100
score and one entry per check with a status and an actionable message.



# Selects a hook from ideated content and generates full content.
Source: https://docs.parallellabs.app/api-reference/content/selects-a-hook-from-ideated-content-and-generates-full-content

/openapi.json post /api/v0/content/{contentId}/select-hook



# Start ICP generation in the background. Returns immediately.
Source: https://docs.parallellabs.app/api-reference/content/start-icp-generation-in-the-background-returns-immediately

/openapi.json post /api/v0/content/profiles/generate



# Trigger a content campaign to generate content.
Source: https://docs.parallellabs.app/api-reference/content/trigger-a-content-campaign-to-generate-content

/openapi.json post /api/v0/content/campaigns/{contentCampaignId}/trigger



# Update a canvas design's name, size, or animation settings.
Source: https://docs.parallellabs.app/api-reference/content/update-a-canvas-designs-name-size-or-animation-settings

/openapi.json patch /api/v0/content/canvases/{canvas_id}



# Update the assignment status of a content item (for reviewers).
Source: https://docs.parallellabs.app/api-reference/content/update-the-assignment-status-of-a-content-item-for-reviewers

/openapi.json put /api/v0/content/{contentId}/assignment-status



# Update the scheduled publish date and metadata for already-scheduled content.
Source: https://docs.parallellabs.app/api-reference/content/update-the-scheduled-publish-date-and-metadata-for-already-scheduled-content

/openapi.json put /api/v0/content/{contentId}/schedule



# Update the status of a content item.
Source: https://docs.parallellabs.app/api-reference/content/update-the-status-of-a-content-item

/openapi.json put /api/v0/content/{contentId}/status



# Upload an image to attach as a design reference in the AI chat.
Source: https://docs.parallellabs.app/api-reference/content/upload-an-image-to-attach-as-a-design-reference-in-the-ai-chat

/openapi.json post /api/v0/content/canvases/{canvas_id}/upload



# Assign departments to a company user (member).
Source: https://docs.parallellabs.app/api-reference/departments/assign-departments-to-a-company-user-member

/openapi.json put /api/v0/company-users/{company_user_id}/departments



# Assign departments to an employee.
Source: https://docs.parallellabs.app/api-reference/departments/assign-departments-to-an-employee

/openapi.json put /api/v0/employees/{employee_id}/departments



# Create a new department.
Source: https://docs.parallellabs.app/api-reference/departments/create-a-new-department

/openapi.json post /api/v0/departments



# Delete a department.
Source: https://docs.parallellabs.app/api-reference/departments/delete-a-department

/openapi.json delete /api/v0/departments/{department_id}



# Get all departments for a company.
Source: https://docs.parallellabs.app/api-reference/departments/get-all-departments-for-a-company

/openapi.json get /api/v0/departments



# Get departments assigned to a company user.
Source: https://docs.parallellabs.app/api-reference/departments/get-departments-assigned-to-a-company-user

/openapi.json get /api/v0/company-users/{company_user_id}/departments



# Get departments assigned to an employee.
Source: https://docs.parallellabs.app/api-reference/departments/get-departments-assigned-to-an-employee

/openapi.json get /api/v0/employees/{employee_id}/departments



# Update a department.
Source: https://docs.parallellabs.app/api-reference/departments/update-a-department

/openapi.json put /api/v0/departments/{department_id}



# Get a documentation page by ID.
Source: https://docs.parallellabs.app/api-reference/documentation/get-a-documentation-page-by-id

/openapi.json get /api/v0/documentation/{pageId}



# Keyword search over documentation pages.
Source: https://docs.parallellabs.app/api-reference/documentation/keyword-search-over-documentation-pages

/openapi.json get /api/v0/documentation/search
Matches are ranked by how many query tokens appear in the title (weighted)
and content, and each result includes short snippets around the matches
so the caller can decide whether to fetch the full page via
GET /api/v0/documentation/{id}.



# List documentation pages.
Source: https://docs.parallellabs.app/api-reference/documentation/list-documentation-pages

/openapi.json get /api/v0/documentation
Omit ``parentId`` to list top-level categories, or pass a parent page ID
to list its direct children. Responses exclude the full page content so
payloads stay small — fetch a specific page with
GET /api/v0/documentation/{id} when you need the body.



# List top-level documentation categories (pages with no parent).
Source: https://docs.parallellabs.app/api-reference/documentation/list-top-level-documentation-categories-pages-with-no-parent

/openapi.json get /api/v0/documentation/categories
Use this first to see what areas of documentation exist, then call
GET /api/v0/documentation?parentId={id} to drill into one.



# Submit a request for new or improved documentation.
Source: https://docs.parallellabs.app/api-reference/documentation/submit-a-request-for-new-or-improved-documentation

/openapi.json post /api/v0/documentation/requests
Creates a support ticket of type "feature_request" tagged as a
documentation request so the docs team can triage it alongside
other feedback.



# Convert uploaded files to text format.
Source: https://docs.parallellabs.app/api-reference/documents/convert-uploaded-files-to-text-format

/openapi.json post /api/v0/documents/convert
Extracts text content from uploaded files without saving them to the knowledge base.
Optionally generates summaries for large documents.



# Copy a document to a different folder.
Source: https://docs.parallellabs.app/api-reference/documents/copy-a-document-to-a-different-folder

/openapi.json post /api/v0/documents/copy
Creates a copy of the document in the specified target folder.



# Create a new document (API-only).
Source: https://docs.parallellabs.app/api-reference/documents/create-a-new-document-api-only

/openapi.json post /api/v0/documents/create
Supports two modes:
1. JSON body for text content: { name, content, path, tags }
2. Multipart form for file upload: file, path, tags

CompanyId is derived from the API key.



# Create a new folder in the knowledge base.
Source: https://docs.parallellabs.app/api-reference/documents/create-a-new-folder-in-the-knowledge-base

/openapi.json post /api/v0/documents/folders
Creates a folder at the specified path.



# Create or update a document.
Source: https://docs.parallellabs.app/api-reference/documents/create-or-update-a-document

/openapi.json post /api/v0/documents
Creates a new document or updates an existing document with the provided content.



# Delete a document from the knowledge base.
Source: https://docs.parallellabs.app/api-reference/documents/delete-a-document-from-the-knowledge-base

/openapi.json delete /api/v0/documents/{documentId}
Permanently removes the document and all associated embeddings.



# Delete a folder from the knowledge base.
Source: https://docs.parallellabs.app/api-reference/documents/delete-a-folder-from-the-knowledge-base

/openapi.json delete /api/v0/documents/folders/{folderId}
If delete=true, also deletes all documents inside the folder.
Otherwise, documents are moved to the root folder.



# Get all folders in the knowledge base.
Source: https://docs.parallellabs.app/api-reference/documents/get-all-folders-in-the-knowledge-base

/openapi.json get /api/v0/documents/folders
Returns the folder structure with optional tag information.



# Get document content by ID (API-only).
Source: https://docs.parallellabs.app/api-reference/documents/get-document-content-by-id-api-only

/openapi.json get /api/v0/documents/{documentId}/content
Returns the full text content of the document.
CompanyId is derived from the API key.



# Get document metadata by ID.
Source: https://docs.parallellabs.app/api-reference/documents/get-document-metadata-by-id

/openapi.json get /api/v0/documents/{companyId}/document/{documentId}
Returns document details including name, path, type, and tags.



# Get document metadata by ID (API-only).
Source: https://docs.parallellabs.app/api-reference/documents/get-document-metadata-by-id-api-only

/openapi.json get /api/v0/documents/{documentId}/details
Returns document details including name, path, type, and tags.
CompanyId is derived from the API key.



# Get recently deleted documents.
Source: https://docs.parallellabs.app/api-reference/documents/get-recently-deleted-documents

/openapi.json get /api/v0/documents/recently-deleted
Returns documents deleted within the specified number of days.



# Get the text content of a document.
Source: https://docs.parallellabs.app/api-reference/documents/get-the-text-content-of-a-document

/openapi.json get /api/v0/documents/{companyId}/{documentId}
Returns the full text content extracted from the document.



# List documents and folders at a path.
Source: https://docs.parallellabs.app/api-reference/documents/list-documents-and-folders-at-a-path

/openapi.json get /api/v0/documents
Returns all documents and folders at the specified path,
sorted with folders first, then files alphabetically.



# Move a folder to a different parent folder.
Source: https://docs.parallellabs.app/api-reference/documents/move-a-folder-to-a-different-parent-folder

/openapi.json put /api/v0/documents/folders/{folderId}/move
Updates the folder location and propagates the change to all child documents.



# Move one or more documents to a different folder.
Source: https://docs.parallellabs.app/api-reference/documents/move-one-or-more-documents-to-a-different-folder

/openapi.json post /api/v0/documents/move
Accepts either a single documentId or a documentIds list and moves
each document to the specified target folder.



# Rename a folder and update all related documents, embeddings, and shares.
Source: https://docs.parallellabs.app/api-reference/documents/rename-a-folder-and-update-all-related-documents-embeddings-and-shares

/openapi.json put /api/v0/documents/folders/{folderId}/rename
Updates the folder name and propagates the change to all child documents.



# Restore a deleted document.
Source: https://docs.parallellabs.app/api-reference/documents/restore-a-deleted-document

/openapi.json post /api/v0/documents/{documentId}/restore
Restores a previously deleted document from the recycle bin.



# Search documents by name, content (vector search), and tags.
Source: https://docs.parallellabs.app/api-reference/documents/search-documents-by-name-content-vector-search-and-tags

/openapi.json get /api/v0/documents/search
Performs semantic search across the knowledge base.



# Search documents using POST (API-only, for n8n compatibility).
Source: https://docs.parallellabs.app/api-reference/documents/search-documents-using-post-api-only-for-n8n-compatibility

/openapi.json post /api/v0/documents/search
Performs semantic search across the knowledge base.
CompanyId is derived from the API key.



# Update an existing document (API-only).
Source: https://docs.parallellabs.app/api-reference/documents/update-an-existing-document-api-only

/openapi.json put /api/v0/documents/{documentId}/update
Updates the document content. Optionally update name and tags.
CompanyId is derived from the API key.



# Update tags for a folder.
Source: https://docs.parallellabs.app/api-reference/documents/update-tags-for-a-folder

/openapi.json put /api/v0/documents/folders/{folderId}/tags
Sets the auto-apply tags for documents created in this folder.



# Upload a document file to the knowledge base.
Source: https://docs.parallellabs.app/api-reference/documents/upload-a-document-file-to-the-knowledge-base

/openapi.json post /api/v0/documents/upload
Supports PDF, DOCX, TXT, and other text-based file formats.
Premium PDF processing is available for better text extraction.



# Check a root-domain registration's availability and price.
Source: https://docs.parallellabs.app/api-reference/domains/check-a-root-domain-registrations-availability-and-price

/openapi.json post /api/v0/domains/check-availability
This read-only endpoint is only for a domain registration purchase. It does
not connect a domain the customer already owns; use domain-connections for
SSL termination and routing instead.



# Connect a customer-owned domain or subdomain for SSL termination.
Source: https://docs.parallellabs.app/api-reference/domains/connect-a-customer-owned-domain-or-subdomain-for-ssl-termination

/openapi.json post /api/v0/domain-connections
This consumes one branded-domain connection slot. Additional slots are a
$5/month add-on. It does not purchase, register, or move DNS for the domain;
create the returned A record at the domain's existing DNS host.



# Get the current one-time price to register one root domain.
Source: https://docs.parallellabs.app/api-reference/domains/get-the-current-one-time-price-to-register-one-root-domain

/openapi.json get /api/v0/domains/pricing
The returned customerPrice applies the same request-scoped whitelabel add-on
multiplier as mailbox orders. It is not the $5/month branded-domain
connection add-on price. Use it before asking the user to approve a purchase.



# List branded-domain connections and the connection-slot limit.
Source: https://docs.parallellabs.app/api-reference/domains/list-branded-domain-connections-and-the-connection-slot-limit

/openapi.json get /api/v0/domain-connections
A connection attaches a domain or subdomain the customer already owns to
SSL termination and routing for tracking links, landing pages, and published
websites. It does not purchase or register a domain.



# List DNS records for a domain registered through this platform.
Source: https://docs.parallellabs.app/api-reference/domains/list-dns-records-for-a-domain-registered-through-this-platform

/openapi.json get /api/v0/domains/{domain_id}/dns
Customer-owned domains connected for SSL termination keep DNS at their
existing DNS host. Read records before proposing any change.



# Purchase a root-domain registration and create its pending connection.
Source: https://docs.parallellabs.app/api-reference/domains/purchase-a-root-domain-registration-and-create-its-pending-connection

/openapi.json post /api/v0/domains/purchase
This is a financial, non-idempotent action. First call the pricing endpoint,
obtain approval, then set purchaseConfirmed to true. Never blindly retry a
timeout; check the domain list first to avoid a duplicate purchase. This is
separate from connecting a customer-owned domain for SSL termination and routing.



# Update DNS records for a domain registered through this platform.
Source: https://docs.parallellabs.app/api-reference/domains/update-dns-records-for-a-domain-registered-through-this-platform

/openapi.json put /api/v0/domains/{domain_id}/dns
This is an irreversible infrastructure change. Read the current records,
preserve records you are not intentionally changing, obtain approval, and
set dnsChangeConfirmed to true. Never blindly retry a timeout.



# Chat with an employee.
Source: https://docs.parallellabs.app/api-reference/employees/chat-with-an-employee

/openapi.json post /api/v0/employees/chat
Send a message to a specific AI employee and receive a response based on
their persona and your company's knowledge base. A new chat is created when
chatId is omitted; pass the returned chatId back to continue the conversation.



# Create a new employee from a persona template.
Source: https://docs.parallellabs.app/api-reference/employees/create-a-new-employee-from-a-persona-template

/openapi.json post /api/v0/employees/create-from-persona
Takes a persona ID and creates an employee with the persona's attributes.
Optionally allows overriding specific fields.



# Create one or more AI employees.
Source: https://docs.parallellabs.app/api-reference/employees/create-one-or-more-ai-employees

/openapi.json post /api/v0/employees



# Delete an employee.
Source: https://docs.parallellabs.app/api-reference/employees/delete-an-employee

/openapi.json delete /api/v0/employees/{employeeId}



# Get a single employee persona by ID.
Source: https://docs.parallellabs.app/api-reference/employees/get-a-single-employee-persona-by-id

/openapi.json get /api/v0/employee-personas/{personaId}



# Get a specific employee by ID.
Source: https://docs.parallellabs.app/api-reference/employees/get-a-specific-employee-by-id

/openapi.json get /api/v0/employees/{employeeId}



# Get all employee persona templates.
Source: https://docs.parallellabs.app/api-reference/employees/get-all-employee-persona-templates

/openapi.json get /api/v0/employee-personas



# Get all employees for a company.
Source: https://docs.parallellabs.app/api-reference/employees/get-all-employees-for-a-company

/openapi.json get /api/v0/companies/{companyId}/employees



# Get all unique persona categories.
Source: https://docs.parallellabs.app/api-reference/employees/get-all-unique-persona-categories

/openapi.json get /api/v0/employee-personas/categories



# Get the result of an async chat request.
Source: https://docs.parallellabs.app/api-reference/employees/get-the-result-of-an-async-chat-request

/openapi.json get /api/v0/employees/chat/async/{chat_id}
Poll this endpoint with the chat ID returned from the async chat endpoint.



# List all employees (API-only, companyId from API key).
Source: https://docs.parallellabs.app/api-reference/employees/list-all-employees-api-only-companyid-from-api-key

/openapi.json get /api/v0/employees
Returns employees for the company associated with the API key.



# Search employee personas.
Source: https://docs.parallellabs.app/api-reference/employees/search-employee-personas

/openapi.json get /api/v0/employee-personas/search



# Start an async chat with an employee.
Source: https://docs.parallellabs.app/api-reference/employees/start-an-async-chat-with-an-employee

/openapi.json post /api/v0/employees/chat/async
Returns immediately with a chat ID that can be used to poll for the response.



# Update an employee.
Source: https://docs.parallellabs.app/api-reference/employees/update-an-employee

/openapi.json post /api/v0/employees/{employeeId}



# Apply a template overlay to an existing image.
Source: https://docs.parallellabs.app/api-reference/images/apply-a-template-overlay-to-an-existing-image

/openapi.json post /api/v0/images/apply-template
Takes a source image and applies a template overlay from a brand image,
then saves and returns the resulting image.



# Delete an image by ID.
Source: https://docs.parallellabs.app/api-reference/images/delete-an-image-by-id

/openapi.json delete /api/v0/images/{imageId}
Permanently removes the image and any associated data.



# Download a stock image and save to S3.
Source: https://docs.parallellabs.app/api-reference/images/download-a-stock-image-and-save-to-s3

/openapi.json post /api/v0/stock-images/download



# Generate an AI image from a text prompt.
Source: https://docs.parallellabs.app/api-reference/images/generate-an-ai-image-from-a-text-prompt

/openapi.json post /api/v0/images
Creates an image using the specified model (GPT Image or Google) and
returns its ID and public URL when generation completes.



# Get an image by ID.
Source: https://docs.parallellabs.app/api-reference/images/get-an-image-by-id

/openapi.json get /api/v0/images/{imageId}
Returns the image metadata including URLs and generation details.



# Get list of available stock image providers.
Source: https://docs.parallellabs.app/api-reference/images/get-list-of-available-stock-image-providers

/openapi.json get /api/v0/stock-images/providers



# List all images for a company.
Source: https://docs.parallellabs.app/api-reference/images/list-all-images-for-a-company

/openapi.json get /api/v0/images
Returns a paginated list of images with their metadata.



# List all supported image models.
Source: https://docs.parallellabs.app/api-reference/images/list-all-supported-image-models

/openapi.json get /api/v0/images/models
Returns a list of available image generation models with their
IDs, names, and providers. This endpoint is public and does not
require authentication.



# Search for stock images from Unsplash and/or Pexels.
Source: https://docs.parallellabs.app/api-reference/images/search-for-stock-images-from-unsplash-andor-pexels

/openapi.json get /api/v0/stock-images/search



# Upload an image file directly.
Source: https://docs.parallellabs.app/api-reference/images/upload-an-image-file-directly

/openapi.json post /api/v0/images/upload
Accepts image files (PNG, JPG, JPEG, GIF, WEBP) up to 10MB.
The original image is stored as-is; format conversion happens at
serve time via the ?format= query parameter.



# Bulk import IMAP inboxes from CSV data (Instantly.ai compliant format).
Source: https://docs.parallellabs.app/api-reference/inbox/bulk-import-imap-inboxes-from-csv-data-instantlyai-compliant-format

/openapi.json post /api/v0/inboxes/bulk-import
Creates integrations, inbox records, and sender profiles for each inbox.
Optionally enables warmup for imported inboxes.



# Create a new inbox.
Source: https://docs.parallellabs.app/api-reference/inbox/create-a-new-inbox

/openapi.json post /api/v0/inboxes



# Delete a draft message.
Source: https://docs.parallellabs.app/api-reference/inbox/delete-a-draft-message

/openapi.json delete /api/v0/inboxes/{inboxId}/drafts/{draftId}



# Delete an inbox.
Source: https://docs.parallellabs.app/api-reference/inbox/delete-an-inbox

/openapi.json delete /api/v0/inboxes/{inboxId}



# Disable warmup for an IMAP inbox.
Source: https://docs.parallellabs.app/api-reference/inbox/disable-warmup-for-an-imap-inbox

/openapi.json delete /api/v0/inboxes/{inboxId}/warmup



# Edit an existing reply draft using AI based on user instructions.
Source: https://docs.parallellabs.app/api-reference/inbox/edit-an-existing-reply-draft-using-ai-based-on-user-instructions

/openapi.json post /api/v0/inbox/messages/{messageId}/edit-reply



# Enable warmup for an IMAP inbox.
Source: https://docs.parallellabs.app/api-reference/inbox/enable-warmup-for-an-imap-inbox

/openapi.json post /api/v0/inboxes/{inboxId}/warmup



# Generate AI content for a new email composition.
Source: https://docs.parallellabs.app/api-reference/inbox/generate-ai-content-for-a-new-email-composition

/openapi.json post /api/v0/inbox/compose/generate



# Generate an AI reply to a message using the shared email prompt system.
Source: https://docs.parallellabs.app/api-reference/inbox/generate-an-ai-reply-to-a-message-using-the-shared-email-prompt-system

/openapi.json post /api/v0/inbox/messages/{messageId}/generate-reply



# Get all draft messages for an inbox.
Source: https://docs.parallellabs.app/api-reference/inbox/get-all-draft-messages-for-an-inbox

/openapi.json get /api/v0/inboxes/{inboxId}/drafts



# Get full message details including body.
Source: https://docs.parallellabs.app/api-reference/inbox/get-full-message-details-including-body

/openapi.json get /api/v0/inboxes/messages/{messageId}



# Get messages for a specific inbox.
Source: https://docs.parallellabs.app/api-reference/inbox/get-messages-for-a-specific-inbox

/openapi.json get /api/v0/inboxes/{inboxId}/messages



# Get messages from all inboxes or filtered by inbox IDs.
Source: https://docs.parallellabs.app/api-reference/inbox/get-messages-from-all-inboxes-or-filtered-by-inbox-ids

/openapi.json get /api/v0/inboxes/messages
Supports filtering by message types (comma-separated) and inbox IDs.



# Get overall statistics for inboxes.
Source: https://docs.parallellabs.app/api-reference/inbox/get-overall-statistics-for-inboxes

/openapi.json get /api/v0/inboxes/statistics



# Get warmup analytics for enabled inboxes.
Source: https://docs.parallellabs.app/api-reference/inbox/get-warmup-analytics-for-enabled-inboxes

/openapi.json post /api/v0/inboxes/warmup/analytics



# List all IMAP inboxes that can be used for warmup.
Source: https://docs.parallellabs.app/api-reference/inbox/list-all-imap-inboxes-that-can-be-used-for-warmup

/openapi.json get /api/v0/inboxes/warmup/list



# List all inboxes in the company for selection purposes.
Source: https://docs.parallellabs.app/api-reference/inbox/list-all-inboxes-in-the-company-for-selection-purposes

/openapi.json get /api/v0/inboxes/for-selection
Returns all inboxes regardless of sharing status, useful for sender profile dropdowns.



# List inboxes visible to the user.
Source: https://docs.parallellabs.app/api-reference/inbox/list-inboxes-visible-to-the-user

/openapi.json get /api/v0/inboxes
Only returns shared inboxes or private inboxes owned by the user.
Supports optional 'query' parameter to search by name or email.



# Send a draft message.
Source: https://docs.parallellabs.app/api-reference/inbox/send-a-draft-message

/openapi.json post /api/v0/inboxes/{inboxId}/drafts/{draftId}/send



# Send a new composed email via inbox.
Source: https://docs.parallellabs.app/api-reference/inbox/send-a-new-composed-email-via-inbox

/openapi.json post /api/v0/inbox/{inboxId}/compose



# Send a reply to a message.
Source: https://docs.parallellabs.app/api-reference/inbox/send-a-reply-to-a-message

/openapi.json post /api/v0/inbox/messages/{messageId}/reply



# Set a custom warmup filter tag for selected inboxes in a company.
Source: https://docs.parallellabs.app/api-reference/inbox/set-a-custom-warmup-filter-tag-for-selected-inboxes-in-a-company

/openapi.json post /api/v0/inboxes/warmup/filter-tag



# Transfer ownership of an inbox to another user in the company.
Source: https://docs.parallellabs.app/api-reference/inbox/transfer-ownership-of-an-inbox-to-another-user-in-the-company

/openapi.json post /api/v0/inboxes/{inboxId}/transfer



# Update an existing inbox.
Source: https://docs.parallellabs.app/api-reference/inbox/update-an-existing-inbox

/openapi.json put /api/v0/inboxes/{inboxId}



# List enabled CRM integrations, to discover a valid crmIntegrationId.
Source: https://docs.parallellabs.app/api-reference/integrations/list-enabled-crm-integrations-to-discover-a-valid-crmintegrationid

/openapi.json get /api/v0/crm/integrations
Use the returned id when setting a sequence's crmIntegrationId (with
crmSyncEnabled=true). Pass companyId as a query parameter.



# List HeyReach campaigns for an integration, to find a step campaignId.
Source: https://docs.parallellabs.app/api-reference/integrations/list-heyreach-campaigns-for-an-integration-to-find-a-step-campaignid

/openapi.json get /api/v0/integrations/heyreach/campaigns
Pass companyId and the heyreachIntegrationId as query params; use a returned
campaign id as the heyreach_campaign step's campaignId.



# List integrations of a given type, to discover valid integration IDs.
Source: https://docs.parallellabs.app/api-reference/integrations/list-integrations-of-a-given-type-to-discover-valid-integration-ids

/openapi.json get /api/v0/integrations/by-type/{integration_type}
Use this to find IDs for sequence and sender-profile channel fields. Common
integration_type path values: browser_session (with query type=linkedin_login)
for LinkedIn accounts, sms for Twilio/phone, scheduling for calendar booking,
email for mailboxes. Supports comma-separated types (e.g. 'tools,toolset') and
an optional provider query param (e.g. provider=composio) to fetch all
integrations for that provider regardless of type.



# List Meta Ads custom audiences for an integration, to find a step audienceId.
Source: https://docs.parallellabs.app/api-reference/integrations/list-meta-ads-custom-audiences-for-an-integration-to-find-a-step-audienceid

/openapi.json get /api/v0/integrations/meta-ads/{integration_id}/audiences
Pass companyId as a query param and the integration in the path; use a
returned audience id as the meta_ads_audience step's audienceId.

API Reference: https://developers.facebook.com/docs/marketing-api/audiences/guides/custom-audiences



# List Salesforge sequences in a workspace, to find a step sequenceId.
Source: https://docs.parallellabs.app/api-reference/integrations/list-salesforge-sequences-in-a-workspace-to-find-a-step-sequenceid

/openapi.json post /api/v0/integrations/salesforge/sequences
POST companyId, the salesforgeIntegrationId, and a workspaceId; use a
returned sequence id as the salesforge_campaign step's sequenceId.



# List Salesforge workspaces for an integration, to find a step workspaceId.
Source: https://docs.parallellabs.app/api-reference/integrations/list-salesforge-workspaces-for-an-integration-to-find-a-step-workspaceid

/openapi.json post /api/v0/integrations/salesforge/workspaces
POST companyId and the salesforgeIntegrationId; use a returned workspace id
as the salesforge_campaign step's workspaceId.



# List the sending phone numbers available on a Twilio integration.
Source: https://docs.parallellabs.app/api-reference/integrations/list-the-sending-phone-numbers-available-on-a-twilio-integration

/openapi.json get /api/v0/integrations/twilio/{integration_id}/phone-numbers
Use this to pick a number for a sender profile's phone or a sequence's
fallback phoneNumbers. Pass companyId as a query parameter.



# Parallel AI API Reference — Overview & Authentication
Source: https://docs.parallellabs.app/api-reference/introduction

Complete reference for the Parallel AI API. All endpoints are under /api/v0/ and return JSON. Authenticate with a Company API Key or Personal Access Key.

The Parallel AI REST API gives you programmatic access to every resource in your Parallel account — agents, conversations, smart lists, sequences, documents, audio, and more. All endpoints are served from a single base URL, return JSON, and follow consistent pagination and error conventions.

## Base URL

```
https://api.parallellabs.app
```

Every endpoint path begins with `/api/v0/`. For example, to list your agents you call:

```
GET https://api.parallellabs.app/api/v0/agents
```

## Authentication

The API supports two authentication methods.

### Company API Key

Include your Company API Key in the `X-API-Key` request header. This is the simplest method and is scoped to your company account.

```http theme={null}
X-API-Key: YOUR_API_KEY
```

### Personal Access Key (pak\_)

Personal Access Keys begin with `pak_`. Pass the key as a Bearer token in the `Authorization` header. If the key is associated with multiple companies, also include the `X-Company-ID` header to specify which company context to use.

```http theme={null}
Authorization: Bearer pak_YOUR_PERSONAL_ACCESS_KEY
X-Company-ID: company_abc123
```

<Note>
  Keep your API keys secret. Never expose them in client-side code or public repositories. Rotate them immediately if you suspect they have been compromised.
</Note>

## Response Format

All responses — including errors — are JSON. Successful responses return the requested resource or a confirmation message. The HTTP status code indicates whether the request succeeded.

### Error Format

When a request fails, the response body contains a single `error` field describing what went wrong:

```json theme={null}
{
  "error": "Agent not found"
}
```

## HTTP Status Codes

| Status                      | Meaning                                                                                   |
| --------------------------- | ----------------------------------------------------------------------------------------- |
| `200 OK`                    | The request succeeded and the response body contains the result.                          |
| `201 Created`               | A new resource was created successfully.                                                  |
| `400 Bad Request`           | The request body or query parameters were invalid. Check the `error` message for details. |
| `401 Unauthorized`          | Missing or invalid API key. Verify your credentials.                                      |
| `403 Forbidden`             | Your plan or role does not have access to this feature or resource.                       |
| `404 Not Found`             | The requested resource does not exist.                                                    |
| `429 Too Many Requests`     | You have exceeded the rate limit. Wait before retrying.                                   |
| `500 Internal Server Error` | An unexpected error occurred on the server.                                               |

## Pagination

List endpoints accept `page` and `pageSize` query parameters and return a consistent envelope:

| Parameter  | Type    | Default | Description                                 |
| ---------- | ------- | ------- | ------------------------------------------- |
| `page`     | integer | `1`     | The page number to retrieve.                |
| `pageSize` | integer | `10`    | Number of items per page. Maximum is `100`. |

**Paginated response envelope:**

```json theme={null}
{
  "items": [ /* array of resource objects */ ],
  "total": 42,
  "pages": 5,
  "hasNext": true
}
```

| Field     | Type    | Description                                               |
| --------- | ------- | --------------------------------------------------------- |
| `items`   | array   | The list of resource objects for the current page.        |
| `total`   | integer | Total number of matching resources across all pages.      |
| `pages`   | integer | Total number of pages at the current `pageSize`.          |
| `hasNext` | boolean | `true` if there is at least one more page after this one. |

## API Resources

<CardGroup>
  <Card title="Agents" icon="robot" href="/api-reference/agents/list-all-agents-for-the-current-company">
    Create and manage AI agents with custom prompts, channels, and branding.
  </Card>

  <Card title="Chat" icon="messages" href="/api-reference/chat/create-a-chat-completion-openai-compatible">
    Send messages to agents and retrieve conversation histories.
  </Card>

  <Card title="Smart Lists" icon="list-check" href="/api-reference/lists/list-all-smart-lists-for-the-current-company">
    Manage contact lists and the contacts within them.
  </Card>

  <Card title="Sequences" icon="diagram-next" href="/api-reference/sequences/get-sequence-step-with-sender-profile-and-integration-details">
    Build and trigger multi-step automated messaging sequences.
  </Card>

  <Card title="Documents" icon="file-lines" href="/api-reference/documents/list-documents-and-folders-at-a-path">
    Upload and manage knowledge base documents for your agents.
  </Card>

  <Card title="Audio / Media" icon="waveform" href="/api-reference/audio/list-all-audio-files">
    Transcribe audio and manage media assets.
  </Card>

  <Card title="Browser" icon="browser" href="/api-reference/browser/list-available-browser-integrations">
    Extract web content for use in agent knowledge bases.
  </Card>

  <Card title="Keys" icon="key" href="/api-reference/keys/list-all-api-keys-for-the-current-user">
    Create and manage Company API Keys and Personal Access Keys.
  </Card>

  <Card title="MCP" icon="plug" href="/api-reference/mcp/get-mcp-server-information">
    Connect external tools and data sources via the Model Context Protocol.
  </Card>
</CardGroup>


# Create a new API key.
Source: https://docs.parallellabs.app/api-reference/keys/create-a-new-api-key

/openapi.json post /api/v0/api_keys



# Create or get an existing API key.
Source: https://docs.parallellabs.app/api-reference/keys/create-or-get-an-existing-api-key

/openapi.json put /api/v0/api_keys



# Delete an API key.
Source: https://docs.parallellabs.app/api-reference/keys/delete-an-api-key

/openapi.json delete /api/v0/api_keys



# Get a specific API key by company ID and name.
Source: https://docs.parallellabs.app/api-reference/keys/get-a-specific-api-key-by-company-id-and-name

/openapi.json get /api/v0/api_keys/{companyId}/{name}



# List all API keys for the current user.
Source: https://docs.parallellabs.app/api-reference/keys/list-all-api-keys-for-the-current-user

/openapi.json get /api/v0/api_keys



# Ask the thread's in-flight agent run to wrap up (same signal a new
Source: https://docs.parallellabs.app/api-reference/landing-pages/ask-the-threads-in-flight-agent-run-to-wrap-up-same-signal-a-new

/openapi.json post /api/v0/landing-pages/{page_id}/threads/{thread_id}/stop
message sends; the partial reply is already saved).



# Chat endpoint for the AI website builder.
Source: https://docs.parallellabs.app/api-reference/landing-pages/chat-endpoint-for-the-ai-website-builder

/openapi.json post /api/v0/landing-pages/{page_id}/chat
Shares the app builder's agent loop and typed SSE protocol (text /
tool_call / tool_result / files_changed / done / error). The thread is the
source of truth: history is loaded from the landing page's stored threads,
the user message is persisted immediately, and the assistant message
(with its tool-call log) is persisted when the stream ends — the client
never bulk-saves history.

Legacy clients (no `clientMessageId`, inline `messages` history) still get
the old raw-text stream so an older frontend keeps working.



# Create a new chat thread (becomes the active thread).
Source: https://docs.parallellabs.app/api-reference/landing-pages/create-a-new-chat-thread-becomes-the-active-thread

/openapi.json post /api/v0/landing-pages/{page_id}/threads



# Create a new landing page.
Source: https://docs.parallellabs.app/api-reference/landing-pages/create-a-new-landing-page

/openapi.json post /api/v0/landing-pages



# Delete a landing page.
Source: https://docs.parallellabs.app/api-reference/landing-pages/delete-a-landing-page

/openapi.json delete /api/v0/landing-pages/{page_id}



# Delete a thread; returns the refreshed thread view.
Source: https://docs.parallellabs.app/api-reference/landing-pages/delete-a-thread;-returns-the-refreshed-thread-view

/openapi.json delete /api/v0/landing-pages/{page_id}/threads/{thread_id}



# Get a landing page by ID with HTML content.
Source: https://docs.parallellabs.app/api-reference/landing-pages/get-a-landing-page-by-id-with-html-content

/openapi.json get /api/v0/landing-pages/{page_id}



# Get one thread's messages.
Source: https://docs.parallellabs.app/api-reference/landing-pages/get-one-threads-messages

/openapi.json get /api/v0/landing-pages/{page_id}/threads/{thread_id}



# Get paginated list of landing pages.
Source: https://docs.parallellabs.app/api-reference/landing-pages/get-paginated-list-of-landing-pages

/openapi.json get /api/v0/landing-pages



# Get the current preview HTML content from the landing page's project.
Source: https://docs.parallellabs.app/api-reference/landing-pages/get-the-current-preview-html-content-from-the-landing-pages-project

/openapi.json get /api/v0/landing-pages/{page_id}/preview



# List chat threads for a landing page ({threads, activeThreadId}).
Source: https://docs.parallellabs.app/api-reference/landing-pages/list-chat-threads-for-a-landing-page-

/openapi.json get /api/v0/landing-pages/{page_id}/threads



# Publish a landing page to make it publicly accessible.
Source: https://docs.parallellabs.app/api-reference/landing-pages/publish-a-landing-page-to-make-it-publicly-accessible

/openapi.json post /api/v0/landing-pages/{page_id}/publish



# Rename a thread (`title`) or make it active (`active: true`).
Source: https://docs.parallellabs.app/api-reference/landing-pages/rename-a-thread-`title`-or-make-it-active-`active:-true`

/openapi.json patch /api/v0/landing-pages/{page_id}/threads/{thread_id}



# Save chat history for a landing page.
Source: https://docs.parallellabs.app/api-reference/landing-pages/save-chat-history-for-a-landing-page

/openapi.json put /api/v0/landing-pages/{page_id}/chat-history



# Unpublish a landing page to remove public access.
Source: https://docs.parallellabs.app/api-reference/landing-pages/unpublish-a-landing-page-to-remove-public-access

/openapi.json post /api/v0/landing-pages/{page_id}/unpublish



# Update landing page metadata.
Source: https://docs.parallellabs.app/api-reference/landing-pages/update-landing-page-metadata

/openapi.json patch /api/v0/landing-pages/{page_id}



# Update the HTML content of a landing page.
Source: https://docs.parallellabs.app/api-reference/landing-pages/update-the-html-content-of-a-landing-page

/openapi.json put /api/v0/landing-pages/{page_id}/content



# Upload an image to attach as a design reference in the AI chat.
Source: https://docs.parallellabs.app/api-reference/landing-pages/upload-an-image-to-attach-as-a-design-reference-in-the-ai-chat

/openapi.json post /api/v0/landing-pages/{page_id}/upload



# Describe the lead-generation filter surface and list valid enum values
Source: https://docs.parallellabs.app/api-reference/leads/describe-the-lead-generation-filter-surface-and-list-valid-enum-values

/openapi.json get /api/v0/leads/field-options
describe=true returns the schema map — every field the generate endpoint
accepts, its type (enum/text/range/keywords), its group, and where it
lives. Enum option lists are inlined only when scoped with field= or
group=. Call describe=true first (compact), then fetch options per group
as you build filters. Free — no data credits charged.



# Enrich a person or company from any combination of identifiers
Source: https://docs.parallellabs.app/api-reference/leads/enrich-a-person-or-company-from-any-combination-of-identifiers

/openapi.json post /api/v0/leads/enrich
Provide any combination of name, email, phone, company, domain, LinkedIn
URL, or address — more identifiers = more precise matching. Returns full
contact profiles (emails, phones, job title, employer, up to ~75 fields
per match). A no-match returns {results: [], count: 0}, not an error, and
is not charged. Costs 2 data credits per matched call.



# Generate leads by filters, premade segments, and/or buyer-intent keywords
Source: https://docs.parallellabs.app/api-reference/leads/generate-leads-by-filters-premade-segments-andor-buyer-intent-keywords

/openapi.json post /api/v0/leads
At least one of filters / segment / audience is required; combining them
intersects the criteria. Call the field-options endpoint first to learn
the filter surface and valid enum values (never guess enum strings), and
the segments endpoint to find premade segment externalIds. Costs 1 data
credit per call.



# Import leads into a smart list
Source: https://docs.parallellabs.app/api-reference/leads/import-leads-into-a-smart-list

/openapi.json post /api/v0/leads/import
Takes the same criteria as the generate endpoint but, instead of returning
a sample, builds the full audience and writes every matching contact into
the destination smart list as rows. The list must already exist — create
one with the Lists API first and pass its id.

Runs in the background: responds 202 with a taskId as soon as the import is
queued, and rows appear in the list over the next few minutes. Costs 1 data
credit per contact imported. The credit check gates on the FULL match count,
not on `count`, because the whole audience is built regardless of how many
rows are imported — call the generate endpoint first to see totalAvailable.



# Search the premade interest/intent segment taxonomy
Source: https://docs.parallellabs.app/api-reference/leads/search-the-premade-interestintent-segment-taxonomy

/openapi.json get /api/v0/leads/segments
Thousands of B2B/B2C segments — searched, never listed. Returns matching
segments; pass their externalId values (e.g. "b2b_12636") into the
generate endpoint's segment parameter. Free — no data credits charged.



# Add a lookup column to a list.
Source: https://docs.parallellabs.app/api-reference/lists/add-a-lookup-column-to-a-list

/openapi.json post /api/v0/lists/{listId}/lookup
Lookup columns are computed, read-only columns whose values are derived live
(no writes, no stale data) — e.g. each row's sequence status or agent
conversation count. Adding the same lookup type twice is a no-op that
returns the existing column. Discover the available lookup types via
GET /lists/actions (entries that carry a lookupType).



# Add a new column to a list.
Source: https://docs.parallellabs.app/api-reference/lists/add-a-new-column-to-a-list

/openapi.json post /api/v0/lists/{listId}/header



# Add a new empty row to a list.
Source: https://docs.parallellabs.app/api-reference/lists/add-a-new-empty-row-to-a-list

/openapi.json post /api/v0/lists/{listId}/row



# Add multiple rows with data to a list.
Source: https://docs.parallellabs.app/api-reference/lists/add-multiple-rows-with-data-to-a-list

/openapi.json post /api/v0/lists/{listId}/rows
Creates multiple rows with the provided data. Optionally creates new
columns for fields that don't exist and matches existing rows based
on specified fields.

Query parameters:
- createColumns: Create columns that don't exist (default: false)
- matchFields: Comma-separated field names to match existing rows



# Add (or update) a table action on a list.
Source: https://docs.parallellabs.app/api-reference/lists/add-or-update-a-table-action-on-a-list

/openapi.json post /api/v0/lists/{listId}/action
Table actions are automations attached to a list as an action column — e.g.
a research agent that enriches each row, an AI text generator, or an HTTP
request. Set autoRunEnabled to run the action automatically on rows as they
are added.

Discover the available actions and their default parameters via
GET /lists/actions, then pass that template (customized) as actionParameters.
Provide actionId to update an existing action instead of creating one.



# Bulk-update rows, SQL-style: UPDATE <list> SET <set> WHERE <where>.
Source: https://docs.parallellabs.app/api-reference/lists/bulk-update-rows-sql-style:-update-<list>-set-<set>-where-<where>

/openapi.json put /api/v0/lists/{listId}/rows
Apply the same column→value assignments (`set`) to every row matching the
`where` conditions tree in a single call — instead of fetching ids and
looping single-row updates. `where` uses the same conditions shape as action
run-conditions, and may filter on the virtual `rowId` field to target a
single row by id. Updating every row requires an explicit `matchAll: true`
(with no `where`); a missing/empty `where` alone is a 400, never a
match-all, and so is a present-but-malformed `where` or an unknown body
key. Returns how many rows matched and how many actually changed.



# Create a new smart list.
Source: https://docs.parallellabs.app/api-reference/lists/create-a-new-smart-list

/openapi.json post /api/v0/lists
Set listDict.sharing to control who can see the list:
- "Private" (default): only the creator.
- "Shared": anyone in the company.
- "Public": anyone with the list ID, via the unauthenticated
/public/lists/<listId> endpoints.



# Deduplicate rows by the combined value of one or more columns.
Source: https://docs.parallellabs.app/api-reference/lists/deduplicate-rows-by-the-combined-value-of-one-or-more-columns

/openapi.json post /api/v0/lists/{listId}/deduplicate
Rows are grouped by their values in `columnNames`; within each group the
oldest row is kept and the newer duplicates are removed ("delete") or have
their values merged into the kept row first ("merge"). Every name in
`columnNames` must be an existing column on the list — unknown columns
return 400 with the list of valid columns. Runs asynchronously: responds
202 with a taskId.



# Delete a column from a list.
Source: https://docs.parallellabs.app/api-reference/lists/delete-a-column-from-a-list

/openapi.json delete /api/v0/lists/{listId}/header/{headerId}



# Delete a smart list.
Source: https://docs.parallellabs.app/api-reference/lists/delete-a-smart-list

/openapi.json delete /api/v0/lists/{listId}



# Delete a table action (and its action column) from a list.
Source: https://docs.parallellabs.app/api-reference/lists/delete-a-table-action-and-its-action-column-from-a-list

/openapi.json delete /api/v0/lists/{listId}/action/{actionId}



# Delete multiple rows from a list.
Source: https://docs.parallellabs.app/api-reference/lists/delete-multiple-rows-from-a-list

/openapi.json post /api/v0/lists/{listId}/delete



# Duplicate a smart list and all its active row data.
Source: https://docs.parallellabs.app/api-reference/lists/duplicate-a-smart-list-and-all-its-active-row-data

/openapi.json post /api/v0/lists/{listId}/duplicate
Copies every active row, column definition, cell value, cell research and
timestamp, and table action configuration/results. New list, header, and
row IDs make the copy independent. Deleted rows/columns are excluded.
The caller owns the copy; sharing and department restrictions are retained.
Copied automations start disabled. Running jobs, execution history, campaign
recipients, and sequence memberships remain attached to the original list.
Computed lookup columns recalculate for the new list. The entire copy is
committed atomically; an error leaves no partial list.



# Export a list to CSV.
Source: https://docs.parallellabs.app/api-reference/lists/export-a-list-to-csv

/openapi.json get /api/v0/lists/{listId}/export
For lists under 10k rows, returns CSV directly.
For larger lists, creates a background task and emails the download link.



# Fetch one saved lead search, including the results it returned.
Source: https://docs.parallellabs.app/api-reference/lists/fetch-one-saved-lead-search-including-the-results-it-returned

/openapi.json get /api/v0/lead-searches/{searchId}
The results are returned exactly as they were when the search ran, so they
can be reviewed and added to a list without paying to run the search again.



# Get a list's metadata and column headers without any row data.
Source: https://docs.parallellabs.app/api-reference/lists/get-a-lists-metadata-and-column-headers-without-any-row-data

/openapi.json get /api/v0/lists/{listId}/schema
Returns the list's name, type, and ordered column definitions (headers).
Use this to discover the exact column names and types before reading or
writing rows — it avoids fetching the (potentially large) rows payload that
GET /lists/<listId> returns.



# Get a page of a list's rows without the headers or list metadata.
Source: https://docs.parallellabs.app/api-reference/lists/get-a-page-of-a-lists-rows-without-the-headers-or-list-metadata

/openapi.json get /api/v0/lists/{listId}/rows
Returns only the paginated rows payload. Pair this with
GET /lists/<listId>/schema (column definitions) to read a list without the
bundled GET /lists/<listId> call. Supports the same pagination, sorting,
search, and filterModel parameters as the list-detail endpoint.



# Get a public smart list's metadata and column definitions (no auth required).
Source: https://docs.parallellabs.app/api-reference/lists/get-a-public-smart-lists-metadata-and-column-definitions-no-auth-required

/openapi.json get /api/v0/public/lists/{listId}



# Get a specific list with its rows and column headers.
Source: https://docs.parallellabs.app/api-reference/lists/get-a-specific-list-with-its-rows-and-column-headers

/openapi.json get /api/v0/lists/{listId}



# Get paginated rows from a public smart list (no auth required).
Source: https://docs.parallellabs.app/api-reference/lists/get-paginated-rows-from-a-public-smart-list-no-auth-required

/openapi.json get /api/v0/public/lists/{listId}/rows



# Get rows added to a list since a given timestamp.
Source: https://docs.parallellabs.app/api-reference/lists/get-rows-added-to-a-list-since-a-given-timestamp

/openapi.json get /api/v0/lists/{listId}/rows/new
Returns all rows that were created after the specified ISO timestamp.
Useful for incremental synchronization.



# Get the change history for a single row.
Source: https://docs.parallellabs.app/api-reference/lists/get-the-change-history-for-a-single-row

/openapi.json get /api/v0/lists/{listId}/row/{rowId}/history



# Get the queue-processing status for a table action.
Source: https://docs.parallellabs.app/api-reference/lists/get-the-queue-processing-status-for-a-table-action

/openapi.json get /api/v0/lists/{listId}/action/{actionId}/queue-status
Returns aggregate counts of how many rows are pending / processing /
completed / failed for this action. Set includeDetails=true to also get
per-item details. Useful for polling progress after running an action.



# Get the research summary and steps produced by an agent/browser action.
Source: https://docs.parallellabs.app/api-reference/lists/get-the-research-summary-and-steps-produced-by-an-agentbrowser-action

/openapi.json get /api/v0/lists/{listId}/action/{actionId}/summary



# Get the status of a background task by its taskId.
Source: https://docs.parallellabs.app/api-reference/lists/get-the-status-of-a-background-task-by-its-taskid

/openapi.json get /api/v0/lists/tasks/{taskId}
Use this to poll an asynchronous operation (e.g. deduplication) that
returned a taskId. Re-fetch until `status` is terminal — "completed",
"failed", or "cancelled" ("pending" and "running" mean keep polling.)



# Import a saved search's results into a smart list — free.
Source: https://docs.parallellabs.app/api-reference/lists/import-a-saved-searchs-results-into-a-smart-list-—-free

/openapi.json post /api/v0/lead-searches/{searchId}/import
Works for any saved search, whichever surface produced it (a Sales Navigator
search or an enrichment lookup). The results were stored when the search ran
and were paid for then, so this reuses them and **charges nothing** rather
than asking the vendor for the same leads twice.

People already on the list are matched on linkedinUrl and enriched in place —
only their empty fields are filled — so importing the same search twice does
not create duplicates. Missing columns are created automatically.

Pass `rowIndexes` to import only some of the results (0-based, indexing into
the search's stored `results`); omit it to import all of them.



# List all smart lists for the current company.
Source: https://docs.parallellabs.app/api-reference/lists/list-all-smart-lists-for-the-current-company

/openapi.json get /api/v0/lists



# List saved lead searches for the company, newest first.
Source: https://docs.parallellabs.app/api-reference/lists/list-saved-lead-searches-for-the-company-newest-first

/openapi.json get /api/v0/lead-searches
Result rows are omitted here — the listing carries the label, result count
and cost so a history panel can render without pulling every stored row.
Fetch a single search by id to get its results back.

Filter by `source` ("sales_navigator" or "enrichment") to show one surface's
history.



# List the available smart-list table actions (the action catalog).
Source: https://docs.parallellabs.app/api-reference/lists/list-the-available-smart-list-table-actions-the-action-catalog

/openapi.json get /api/v0/lists/actions
Returns every action (and lookup column) that can be attached to a list —
its title, description, icon, and a template of default parameters. This is
the source of truth for the "Add action" menu and for AI agents deciding
which action to configure on a column.

Audience actions (Audience Enrichment, Find Leads at Company) are omitted
when the company does not have the Audience Builder feature enabled.



# Move multiple rows from a list to a sequence.
Source: https://docs.parallellabs.app/api-reference/lists/move-multiple-rows-from-a-list-to-a-sequence

/openapi.json post /api/v0/lists/{listId}/move-to-sequence
Adds the specified rows to a sequence as new members. Optionally
removes them from the source list and applies custom field mappings.



# Move multiple rows from a list to another list.
Source: https://docs.parallellabs.app/api-reference/lists/move-multiple-rows-from-a-list-to-another-list

/openapi.json post /api/v0/lists/{listId}/move-to-list
Copies the specified rows to a target list. Optionally removes
them from the source list.



# Remove a saved lead search from the history.
Source: https://docs.parallellabs.app/api-reference/lists/remove-a-saved-lead-search-from-the-history

/openapi.json delete /api/v0/lead-searches/{searchId}
Soft delete — the stored results are kept, so dismissing a search from the
panel never destroys data the company paid for.



# Run a table action against some or all rows of a list.
Source: https://docs.parallellabs.app/api-reference/lists/run-a-table-action-against-some-or-all-rows-of-a-list

/openapi.json post /api/v0/lists/{listId}/action/run
Pass selectedRows to run on specific rows, or omit it to run on every row.
The action runs asynchronously — poll GET /lists/<listId>/action/<actionId>/queue-status
(or re-read the rows) to observe progress.



# Start a Sales Navigator search. Returns immediately with a search id.
Source: https://docs.parallellabs.app/api-reference/lists/start-a-sales-navigator-search-returns-immediately-with-a-search-id

/openapi.json post /api/v0/lead-searches/sales-navigator
**The search runs in the background — this does NOT return the leads.** It
responds 202 with a `searchId`; poll `GET /lead-searches/{searchId}` until
its `status` is no longer "pending", then read the leads from `results`.
A search takes minutes, because each matching profile is enriched
individually upstream.

**Do not retry this endpoint while a search is pending.** Every call starts
a new search and charges for it again; polling the id you already have is
free. If a call appears to time out, the search is still running — find it
in `GET /lead-searches?source=sales_navigator` rather than re-running it.

Nothing is written to a smart list: the caller reviews the results and adds
the ones it wants through `POST /lists/{listId}/rows`, which is what the
dashboard does once the user has picked rows.

The search is specified either by a Sales Navigator people-search URL
(`searchMode` "url") or by structured filters — jobTitle, location,
currentCompany, school, keywords (`searchMode` "filters"). Each result is a
lead row (firstName, lastName, linkedinUrl, headline, positionTitle,
companyName, location, …) ready to be added to a list as-is.

No LinkedIn account or session cookie is required: the search is resolved
from public profile data.

Costs 1 data credit per lead returned, charged when the search finishes —
the same balance enrichment and People Search draw from. A search that
matches nobody or fails is free, and principals with Audience Builder are
not charged at all. Credits are checked against the requested ceiling before
the search starts.



# Start importing the people who engaged with a LinkedIn post into a smart list.
Source: https://docs.parallellabs.app/api-reference/lists/start-importing-the-people-who-engaged-with-a-linkedin-post-into-a-smart-list

/openapi.json post /api/v0/lists/{listId}/import/linkedin-post
Returns immediately with a background task ID; the scrape and import run
asynchronously. Poll `GET /lists/{listId}/tasks` until the task reaches a
terminal status — its `result` holds `imported` and the post summary
(authorName, likesCount, commentsCount).

Scrapes the post's commenters and likers and adds each person to the list as
a lead row (firstName, lastName, linkedinUrl, headline, positionTitle,
companyName, plus engagementType, commentText, sourcePostUrl and postAuthor).
Someone who both commented and liked is recorded once with engagementType
"commented_and_liked".

People already on the list are matched on linkedinUrl and enriched in place —
only their empty fields are filled — so re-importing the same post does not
create duplicates.

`postUrl` must be a linkedin.com/posts/… or linkedin.com/feed/update/… URL;
query parameters are stripped.



# Stop a currently-running table action, cancelling its remaining work.
Source: https://docs.parallellabs.app/api-reference/lists/stop-a-currently-running-table-action-cancelling-its-remaining-work

/openapi.json post /api/v0/lists/{listId}/action/stop



# Update a column type in a list.
Source: https://docs.parallellabs.app/api-reference/lists/update-a-column-type-in-a-list

/openapi.json put /api/v0/lists/{listId}/header/{headerId}



# Update an existing row in a list.
Source: https://docs.parallellabs.app/api-reference/lists/update-an-existing-row-in-a-list

/openapi.json put /api/v0/lists/{listId}/row



# Update an existing smart list.
Source: https://docs.parallellabs.app/api-reference/lists/update-an-existing-smart-list

/openapi.json put /api/v0/lists
The request body is the list object — include id plus any fields to change.
Only the list owner can change name or sharing. Set sharing to:
- "Private": only the owner can view or modify.
- "Shared": anyone in the owner's company can view and modify.
- "Public": anyone with the list ID can read it via the unauthenticated
/public/lists/<listId> endpoints.



# Begin sending a draft campaign immediately.
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/begin-sending-a-draft-campaign-immediately

/openapi.json post /api/v0/marketing-campaigns/{campaignId}/send-now
The campaign must have a senderProfileId, listId, subject, and bodyHtml.
Once this returns, the per-minute cron paces the actual sends so delivery
is spread out over time — expect the status to stay in "sending" while the
backlog drains, then flip to "sent" when the list is exhausted.



# Cancel a scheduled campaign so it will not begin sending.
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/cancel-a-scheduled-campaign-so-it-will-not-begin-sending

/openapi.json post /api/v0/marketing-campaigns/{campaignId}/cancel
Only campaigns in "scheduled" status can be cancelled. Campaigns that are
already sending or sent cannot be cancelled.



# Create a marketing campaign template.
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/create-a-marketing-campaign-template

/openapi.json post /api/v0/marketing-campaign-templates
Author the email body as bodyDocument (the visual-builder JSON); the
server renders bodyHtml from it. Sending bodyHtml is rejected with 400.



# Create a new marketing campaign in draft status.
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/create-a-new-marketing-campaign-in-draft-status

/openapi.json post /api/v0/marketing-campaigns
A newly created campaign is always in draft status and does not send anything
until you call /schedule (to queue it for a future time) or /send-now
(to begin sending immediately). You can keep editing the campaign with PATCH
until it enters sending or sent status.

To create a campaign from a template, GET /api/v0/marketing-campaign-templates/{templateId}
and pass its subject, replyTo, bodyHtml, bodyDocument, senderProfileId, listId,
and fieldMappings here.



# Delete a marketing campaign template. Campaigns already created from it are unaffected.
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/delete-a-marketing-campaign-template-campaigns-already-created-from-it-are-unaffected

/openapi.json delete /api/v0/marketing-campaign-templates/{templateId}



# Get a single marketing campaign by ID.
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/get-a-single-marketing-campaign-by-id

/openapi.json get /api/v0/marketing-campaigns/{campaignId}



# Get a single marketing campaign template, including its full HTML body.
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/get-a-single-marketing-campaign-template-including-its-full-html-body

/openapi.json get /api/v0/marketing-campaign-templates/{templateId}



# List a template's saved content versions.
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/list-a-templates-saved-content-versions

/openapi.json get /api/v0/marketing-campaign-templates/{templateId}/versions
A version is captured whenever the visual document (and therefore the
rendered HTML body) changes.



# List marketing campaign templates.
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/list-marketing-campaign-templates

/openapi.json get /api/v0/marketing-campaign-templates
Templates are reusable email designs (subject, visual-builder body
document with its server-rendered HTML, default sender profile, smart
list, and field mappings). To start a campaign from one, pass its id as
templateId when creating a campaign via POST /api/v0/marketing-campaigns.



# List marketing campaigns for the current company.
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/list-marketing-campaigns-for-the-current-company

/openapi.json get /api/v0/marketing-campaigns
Returns all non-deleted campaigns ordered by creation date (newest first).
Optionally filter by status (draft, scheduled, sending, sent, failed, cancelled).



# List the recipients materialized for this campaign, including per-recipient
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/list-the-recipients-materialized-for-this-campaign-including-per-recipient

/openapi.json get /api/v0/marketing-campaigns/{campaignId}/recipients
send status, open count, click count, and error message where applicable.
Recipients are created when a campaign starts sending; for draft or scheduled
campaigns this list will usually be empty.

Optionally pass ?search=<text> to filter the returned items to recipients
whose email address contains the given substring (case-insensitive). The
aggregate metrics object always reflects all recipients for the campaign,
independent of the search filter.



# List the saved content versions for a campaign, newest first. A version is
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/list-the-saved-content-versions-for-a-campaign-newest-first-a-version-is

/openapi.json get /api/v0/marketing-campaigns/{campaignId}/versions
captured each time the campaign's email body changes.



# Restore a campaign's body to a previously saved version. The restore is
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/restore-a-campaigns-body-to-a-previously-saved-version-the-restore-is

/openapi.json post /api/v0/marketing-campaigns/{campaignId}/versions/{versionId}/restore
itself captured as a new version, so it is always reversible. Only allowed
while the campaign is editable (draft, scheduled, cancelled, or failed).



# Restore a previous version of a template's content.
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/restore-a-previous-version-of-a-templates-content

/openapi.json post /api/v0/marketing-campaign-templates/{templateId}/versions/{versionId}/restore
The restore is itself saved as a new version, so it can always be undone.



# Return aggregate counters (total campaigns, active, sent, opened, clicked,
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/return-aggregate-counters-total-campaigns-active-sent-opened-clicked

/openapi.json get /api/v0/marketing-campaigns/statistics
plus open and click rates) across all of the company's marketing campaigns.



# Return per-hour sent, opened, and clicked counts for the last N days
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/return-per-hour-sent-opened-and-clicked-counts-for-the-last-n-days

/openapi.json get /api/v0/marketing-campaigns/timeseries
(default 7, max 90). Optionally restrict to a single campaign with
?campaignId=…



# Schedule a campaign to begin sending at a specific time.
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/schedule-a-campaign-to-begin-sending-at-a-specific-time

/openapi.json post /api/v0/marketing-campaigns/{campaignId}/schedule
The campaign must have a senderProfileId, listId, subject, and bodyHtml.
Scheduling a campaign that is already sending or sent is rejected.



# Send a one-off test email for a campaign to a specific address.
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/send-a-one-off-test-email-for-a-campaign-to-a-specific-address

/openapi.json post /api/v0/marketing-campaigns/{campaignId}/test-send
Uses the campaign's configured sender profile, subject, and body but does
not affect the recipient list, counters, or send history.



# Soft-delete a marketing campaign. The campaign is hidden from lists but
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/soft-delete-a-marketing-campaign-the-campaign-is-hidden-from-lists-but

/openapi.json delete /api/v0/marketing-campaigns/{campaignId}
the underlying send history and recipient rows are retained.



# Soft-delete a single recipient from a campaign. The row is hidden from
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/soft-delete-a-single-recipient-from-a-campaign-the-row-is-hidden-from

/openapi.json delete /api/v0/marketing-campaigns/{campaignId}/recipients/{recipientId}
listings, search, aggregate metrics, and the send pipeline, but is retained
in the database. Open/click pixel hits that arrive after deletion still
update the row but no longer flow through aggregates.



# Update a marketing campaign. Editable only while the campaign is in draft,
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/update-a-marketing-campaign-editable-only-while-the-campaign-is-in-draft

/openapi.json patch /api/v0/marketing-campaigns/{campaignId}
scheduled, cancelled, or failed status. Editing a scheduled campaign drops
it back to draft until you re-schedule it.



# Update a marketing campaign template.
Source: https://docs.parallellabs.app/api-reference/marketing-campaigns/update-a-marketing-campaign-template

/openapi.json patch /api/v0/marketing-campaign-templates/{templateId}
Changing bodyDocument re-renders bodyHtml on the server and saves a new
version in the template's history. An empty bodyDocument is ignored rather
than clearing the design. bodyHtml is not accepted as input (400).



# Execute an MCP tool call.
Source: https://docs.parallellabs.app/api-reference/mcp/execute-an-mcp-tool-call

/openapi.json post /api/v0/mcp/tools/call
Calls the specified tool with the provided arguments and returns the result in MCP format.

**Example - List Sequences:**

```json
{
"name": "platform_list_capabilities",
"arguments": {"category": "Sequences"}
}
```

**Example - Get Action Details:**

```json
{
"name": "platform_get_action_details",
"arguments": {"operationId": "get_sequences.list_sequences"}
}
```

**Example - Execute Action:**

```json
{
"name": "platform_execute_action",
"arguments": {
"operationId": "get_agents.list_agents"
}
}
```

The response follows MCP format with a `content` array containing the result.



# Get MCP server information.
Source: https://docs.parallellabs.app/api-reference/mcp/get-mcp-server-information

/openapi.json get /api/v0/mcp/info
Returns server capabilities, protocol version, and metadata.

**Connecting to this MCP Server:**

**Claude Desktop** - Add to `claude_desktop_config.json`:

```json
{
"mcpServers": {
"platform": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://api.example.com/api/v0/mcp"],
"env": {"API_KEY": "your_api_key"}
}
}
}
```

**Generic MCP Client:**

```bash
npx mcp-remote https://api.example.com/api/v0/mcp --header "Authorization: Bearer YOUR_API_KEY"
```

**Protocol Version:** 2024-11-05



# List available MCP tools.
Source: https://docs.parallellabs.app/api-reference/mcp/list-available-mcp-tools

/openapi.json get /api/v0/mcp/tools/list
Returns all available tool definitions in MCP format, including input schemas.

**Available Tools:**

- `platform_list_capabilities` - Browse API endpoints by category (Agents, Sequences, Lists, Documents, Content, Images, Audio, Video, Browser, etc.)
- `platform_get_action_details` - Get full specification for an endpoint including required parameters and request body format
- `platform_execute_action` - Execute any API endpoint with automatic authentication

**Usage Flow:**

1. Call `platform_list_capabilities` to discover available actions
2. Call `platform_get_action_details` to understand the parameters needed
3. Call `platform_execute_action` to perform the action

**Authentication:**

- **Company API Key**: Scoped to a single company. No company selection needed.
- **Personal Access Key**: Can access multiple companies by specifying X-Company-ID header.

This gives MCP clients (Claude Desktop, Cursor, Windsurf, etc.) full access to the platform API.



# Create a new memory.
Source: https://docs.parallellabs.app/api-reference/memories/create-a-new-memory

/openapi.json post /api/v0/companies/{companyId}/memories



# Delete a memory (soft delete).
Source: https://docs.parallellabs.app/api-reference/memories/delete-a-memory-soft-delete

/openapi.json delete /api/v0/companies/{companyId}/memories/{memoryId}



# Get a specific memory by ID.
Source: https://docs.parallellabs.app/api-reference/memories/get-a-specific-memory-by-id

/openapi.json get /api/v0/companies/{companyId}/memories/{memoryId}



# Get paginated list of memories for the current user.
Source: https://docs.parallellabs.app/api-reference/memories/get-paginated-list-of-memories-for-the-current-user

/openapi.json get /api/v0/companies/{companyId}/memories



# Toggle a memory's active status.
Source: https://docs.parallellabs.app/api-reference/memories/toggle-a-memorys-active-status

/openapi.json patch /api/v0/companies/{companyId}/memories/{memoryId}/toggle



# Update an existing memory.
Source: https://docs.parallellabs.app/api-reference/memories/update-an-existing-memory

/openapi.json put /api/v0/companies/{companyId}/memories/{memoryId}



# List available AI models.
Source: https://docs.parallellabs.app/api-reference/models/list-available-ai-models

/openapi.json get /api/v0/models
Returns a list of AI models available for chat completions.
This is a public endpoint - no authentication required.
Some models may require Zero Data Retention (ZDR) settings based
on company configuration.



# Create a new keyword monitor.
Source: https://docs.parallellabs.app/api-reference/monitors/create-a-new-keyword-monitor

/openapi.json post /api/v0/monitors
Creates a monitor to track specified keywords across social media platforms
like Reddit, Hacker News, LinkedIn, X (Twitter), and Quora. When matching
posts are found, you'll receive notifications and can generate AI-powered
reply suggestions.



# Delete a monitor.
Source: https://docs.parallellabs.app/api-reference/monitors/delete-a-monitor

/openapi.json delete /api/v0/monitors/{monitor_id}
Permanently deletes the specified monitor. Associated mentions are
preserved but will no longer receive updates.



# Dismiss a mention.
Source: https://docs.parallellabs.app/api-reference/monitors/dismiss-a-mention

/openapi.json post /api/v0/mentions/{mention_id}/dismiss
Marks the mention as dismissed, hiding it from the default view.
Dismissed mentions are not deleted and can still be retrieved
by filtering for dismissed status.



# Generate AI reply suggestions for a mention.
Source: https://docs.parallellabs.app/api-reference/monitors/generate-ai-reply-suggestions-for-a-mention

/openapi.json post /api/v0/mentions/{mention_id}/generate-replies
Uses AI to generate reply suggestions in four different styles:
- value-first: Leads with the benefits of your solution
- educational: Informative and helpful tone
- conversational: Friendly and engaging approach
- problem-solver: Addresses specific pain points

All four styles are generated in a single request. Optionally provide
context about your product/service to personalize the replies.



# Get a specific mention by ID.
Source: https://docs.parallellabs.app/api-reference/monitors/get-a-specific-mention-by-id

/openapi.json get /api/v0/mentions/{mention_id}
Returns full details for a mention including the post content, author
information, engagement metrics, and any generated reply suggestions.
Automatically marks 'new' mentions as 'seen' when retrieved.



# Get a specific monitor by ID.
Source: https://docs.parallellabs.app/api-reference/monitors/get-a-specific-monitor-by-id

/openapi.json get /api/v0/monitors/{monitor_id}
Returns the full configuration for a keyword monitor including its
keywords, platforms, notification settings, and run statistics.



# Get aggregated mention statistics.
Source: https://docs.parallellabs.app/api-reference/monitors/get-aggregated-mention-statistics

/openapi.json get /api/v0/mentions/stats
Returns counts of mentions broken down by status (new, seen, dismissed)
and by platform (Reddit, Hacker News, LinkedIn, X, Quora). Optionally
filter to a specific monitor.



# Get daily mention counts over time, broken down by platform.
Source: https://docs.parallellabs.app/api-reference/monitors/get-daily-mention-counts-over-time-broken-down-by-platform

/openapi.json get /api/v0/mentions/stats/timeseries
Returns a dense (zero-filled) series for the total and for each platform
over the last N days (default 30), shaped for a stacked-area chart.
Optionally filter to a specific monitor.



# List all keyword monitors.
Source: https://docs.parallellabs.app/api-reference/monitors/list-all-keyword-monitors

/openapi.json get /api/v0/monitors
Returns all keyword monitors for the current company. Monitors can be
filtered by status (active or paused) and support pagination.



# List mentions discovered by monitors.
Source: https://docs.parallellabs.app/api-reference/monitors/list-mentions-discovered-by-monitors

/openapi.json get /api/v0/mentions
Returns social media mentions found by your keyword monitors. Mentions
can be filtered by monitor, status (new, seen, dismissed), and platform.
Results are sorted by discovery time, newest first.



# Pause a monitor.
Source: https://docs.parallellabs.app/api-reference/monitors/pause-a-monitor

/openapi.json post /api/v0/monitors/{monitor_id}/pause
Temporarily stops the monitor from checking for new mentions. The monitor
can be resumed later without losing any configuration.



# Resume a paused monitor.
Source: https://docs.parallellabs.app/api-reference/monitors/resume-a-paused-monitor

/openapi.json post /api/v0/monitors/{monitor_id}/resume
Reactivates a paused monitor so it will begin checking for new mentions
again on its regular schedule.



# Trigger a manual run of a monitor.
Source: https://docs.parallellabs.app/api-reference/monitors/trigger-a-manual-run-of-a-monitor

/openapi.json post /api/v0/monitors/{monitor_id}/run
Immediately triggers the monitor to check all configured platforms for
new mentions. The run executes asynchronously in the background.



# Update a monitor's settings.
Source: https://docs.parallellabs.app/api-reference/monitors/update-a-monitors-settings

/openapi.json put /api/v0/monitors/{monitor_id}
Updates the monitor configuration. Only include fields that need to be
changed - unspecified fields will retain their current values.



# Add an existing account to an organization by email.
Source: https://docs.parallellabs.app/api-reference/organizations/add-an-existing-account-to-an-organization-by-email

/openapi.json post /api/v0/organizations/{orgId}/members
Caller must be an owner or admin. Only owners may grant the owner role.



# Change a member's role. Only owners may grant or revoke the owner role.
Source: https://docs.parallellabs.app/api-reference/organizations/change-a-members-role-only-owners-may-grant-or-revoke-the-owner-role

/openapi.json put /api/v0/organizations/{orgId}/members/{memberUid}



# Create a new organization owned by the authenticated user.
Source: https://docs.parallellabs.app/api-reference/organizations/create-a-new-organization-owned-by-the-authenticated-user

/openapi.json post /api/v0/organizations
The organization starts on the free tier; the caller is its first owner.



# List an organization's members. Caller must be an owner or admin.
Source: https://docs.parallellabs.app/api-reference/organizations/list-an-organizations-members-caller-must-be-an-owner-or-admin

/openapi.json get /api/v0/organizations/{orgId}/members



# List organizations with an explicit membership or ownership record.
Source: https://docs.parallellabs.app/api-reference/organizations/list-organizations-with-an-explicit-membership-or-ownership-record

/openapi.json get /api/v0/organizations
Each entry includes a `role` indicating the caller's relationship:
- the caller's organization_users role, when they have a membership row
- otherwise "owner": user is `Organization.ownerUid`
- otherwise "admin": user is in `Organization.owners`

Company collaboration does not grant organization membership and must not
add that company's organization to the organization switcher.



# Remove a member. Only owners may remove other owners.
Source: https://docs.parallellabs.app/api-reference/organizations/remove-a-member-only-owners-may-remove-other-owners

/openapi.json delete /api/v0/organizations/{orgId}/members/{memberUid}



# Rename an organization. Caller must be an owner or admin.
Source: https://docs.parallellabs.app/api-reference/organizations/rename-an-organization-caller-must-be-an-owner-or-admin

/openapi.json put /api/v0/organizations/{orgId}



# A bounded account → person → source map for the interactive graph view.
Source: https://docs.parallellabs.app/api-reference/people/a-bounded-account-→-person-→-source-map-for-the-interactive-graph-view

/openapi.json get /api/v0/people/graph



# Append a text activity to a resolved person.
Source: https://docs.parallellabs.app/api-reference/people/append-a-text-activity-to-a-resolved-person

/openapi.json post /api/v0/people/{personId}/activities
When both ``source`` and ``externalId`` are supplied, repeated requests are
idempotent and return the original activity instead of adding a duplicate.



# Delete one authored/imported activity; source events are read-only.
Source: https://docs.parallellabs.app/api-reference/people/delete-one-authoredimported-activity;-source-events-are-read-only

/openapi.json delete /api/v0/people/{personId}/activities/{activityId}



# List every available person activity: engagement events and custom entries.
Source: https://docs.parallellabs.app/api-reference/people/list-every-available-person-activity:-engagement-events-and-custom-entries

/openapi.json get /api/v0/people/{personId}/activities



# List resolved people for the company.
Source: https://docs.parallellabs.app/api-reference/people/list-resolved-people-for-the-company

/openapi.json get /api/v0/people
Each person unifies every touchpoint across smart lists, sequences,
marketing campaigns, agent conversations, and the inbox. Server-side
paginated; `search` matches names, emails, companies, and any normalized
identifier (so "(415) 555-1234" finds a person stored as "+14155551234").



# One person's full profile: every identifier and every touchpoint.
Source: https://docs.parallellabs.app/api-reference/people/one-persons-full-profile:-every-identifier-and-every-touchpoint

/openapi.json get /api/v0/people/{personId}
Touchpoints are grouped by source type and ordered newest first. This is the
view that answers "everywhere this human appears" without checking each
feature by hand.



# Queue the normal identity refresh, including connected CRM sources.
Source: https://docs.parallellabs.app/api-reference/people/queue-the-normal-identity-refresh-including-connected-crm-sources

/openapi.json post /api/v0/people/sync



# Search for people by name, email, or phone across every source in one call.
Source: https://docs.parallellabs.app/api-reference/people/search-for-people-by-name-email-or-phone-across-every-source-in-one-call

/openapi.json get /api/v0/people/search
Resolves matches through the identity graph, so records that belong to the
same human are attributed to one person rather than returned as unrelated
hits. Each per-source match carries the `personId` it resolved to, and the
`people` array holds the resolved profiles.

Falls back to a direct per-source scan for companies that have not yet built
their identity graph (populated by a nightly background job).



# Summary of the company's identity graph, plus its most recent backfill run.
Source: https://docs.parallellabs.app/api-reference/people/summary-of-the-companys-identity-graph-plus-its-most-recent-backfill-run

/openapi.json get /api/v0/people/stats



# Update display and content fields without changing source idempotency.
Source: https://docs.parallellabs.app/api-reference/people/update-display-and-content-fields-without-changing-source-idempotency

/openapi.json patch /api/v0/people/{personId}/activities/{activityId}



# What this person actually did, newest first, to the microsecond.
Source: https://docs.parallellabs.app/api-reference/people/what-this-person-actually-did-newest-first-to-the-microsecond

/openapi.json get /api/v0/people/{personId}/activity
Where the profile's touchpoints say where a person *appears*, this says how
they *behaved*: opens, link clicks, scheduling-page views, bookings, calls
and sequence status changes, merged into one time-ordered stream.

Currently covers sequence engagement. Marketing campaigns record only
first/last open and click counters rather than individual events, and the
website pixel keeps one row per person rather than one per visit, so
neither can contribute a real event stream yet.

Events attributed to scanners and mail-proxy prefetchers (Microsoft Safe
Links, GoogleImageProxy) are excluded by default -- they are not human
actions, and counting them would overstate engagement. `botCount` reports
how many were held back; pass `includeBots=true` to see them, each flagged
with `isBot` and the provider that was identified.



# Attach a project to a chat.
Source: https://docs.parallellabs.app/api-reference/projects/attach-a-project-to-a-chat

/openapi.json post /api/v0/chats/{chat_id}/project
Links a workspace project to a chat for file access.



# Copy a file.
Source: https://docs.parallellabs.app/api-reference/projects/copy-a-file

/openapi.json post /api/v0/projects/{project_id}/files/cp
Creates a copy of a file at the destination path.



# Create a directory.
Source: https://docs.parallellabs.app/api-reference/projects/create-a-directory

/openapi.json post /api/v0/projects/{project_id}/files/mkdir
Creates a new directory at the specified path.



# Create a new file in a project.
Source: https://docs.parallellabs.app/api-reference/projects/create-a-new-file-in-a-project

/openapi.json post /api/v0/projects/{project_id}/files
Creates a file or directory at the specified path.



# Create a new workspace project.
Source: https://docs.parallellabs.app/api-reference/projects/create-a-new-workspace-project

/openapi.json post /api/v0/projects
Creates a project with optional description, AI instructions, and sharing settings.



# Delete a file.
Source: https://docs.parallellabs.app/api-reference/projects/delete-a-file

/openapi.json delete /api/v0/projects/{project_id}/files/{file_id}
Permanently removes a file from the project.



# Delete a workspace project.
Source: https://docs.parallellabs.app/api-reference/projects/delete-a-workspace-project

/openapi.json delete /api/v0/projects/{project_id}
Deletes the project and all associated files.



# Detach a project from a chat.
Source: https://docs.parallellabs.app/api-reference/projects/detach-a-project-from-a-chat

/openapi.json delete /api/v0/chats/{chat_id}/project
Removes the project link from a chat.



# Export a project file to PDF, DOCX, or CSV format.
Source: https://docs.parallellabs.app/api-reference/projects/export-a-project-file-to-pdf-docx-or-csv-format

/openapi.json get /api/v0/projects/{project_id}/files/{file_id}/export
Returns the file in the requested format as a download.



# Get a file by ID.
Source: https://docs.parallellabs.app/api-reference/projects/get-a-file-by-id

/openapi.json get /api/v0/projects/{project_id}/files/{file_id}
Returns file details including content.



# Get a workspace project by ID.
Source: https://docs.parallellabs.app/api-reference/projects/get-a-workspace-project-by-id

/openapi.json get /api/v0/projects/{project_id}
Returns project if found and accessible to the user.



# Get all workspace projects.
Source: https://docs.parallellabs.app/api-reference/projects/get-all-workspace-projects

/openapi.json get /api/v0/projects
Returns paginated list of projects with optional search and sharing filters.



# Get files in a workspace project.
Source: https://docs.parallellabs.app/api-reference/projects/get-files-in-a-workspace-project

/openapi.json get /api/v0/projects/{project_id}/files
Returns list of files in the specified directory path.



# Get the full file tree for a project.
Source: https://docs.parallellabs.app/api-reference/projects/get-the-full-file-tree-for-a-project

/openapi.json get /api/v0/projects/{project_id}/tree
Returns nested tree structure with all files and directories.



# Get the project attached to a chat.
Source: https://docs.parallellabs.app/api-reference/projects/get-the-project-attached-to-a-chat

/openapi.json get /api/v0/chats/{chat_id}/project
Returns the attached project or null if none.



# Move or rename a file.
Source: https://docs.parallellabs.app/api-reference/projects/move-or-rename-a-file

/openapi.json post /api/v0/projects/{project_id}/files/mv
Moves a file from source path to destination path.



# Publish a file to make it publicly accessible.
Source: https://docs.parallellabs.app/api-reference/projects/publish-a-file-to-make-it-publicly-accessible

/openapi.json post /api/v0/projects/{project_id}/files/{file_id}/publish
Returns the file with publicSlug and publicUrl for sharing.



# Unpublish a file to remove public access.
Source: https://docs.parallellabs.app/api-reference/projects/unpublish-a-file-to-remove-public-access

/openapi.json post /api/v0/projects/{project_id}/files/{file_id}/unpublish
Removes the public URL and makes the file private again.



# Update a workspace project.
Source: https://docs.parallellabs.app/api-reference/projects/update-a-workspace-project

/openapi.json patch /api/v0/projects/{project_id}
All fields are optional. Only provided fields will be updated.



# Update file content.
Source: https://docs.parallellabs.app/api-reference/projects/update-file-content

/openapi.json put /api/v0/projects/{project_id}/files/{file_id}
Replaces the content of an existing file.



# Company-wide scheduled task stats plus a page of tasks with run quality.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/company-wide-scheduled-task-stats-plus-a-page-of-tasks-with-run-quality

/openapi.json get /api/v0/scheduled-tasks/summary
Headline stats across the company, then a page of tasks with their run
counts, average quality scores and score trends.

Open to any company member: the in-chat Scheduled dialog renders the same
panel, and members could already list and manage their scheduled tasks
before this endpoint existed. The Scheduled *page* is super-user gated in
the UI, but that gate is about surfacing a new nav destination, not about
protecting this data.



# Create a scheduled task for an AI employee.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/create-a-scheduled-task-for-an-ai-employee

/openapi.json post /api/v0/scheduled-tasks
scheduleType is one_time (needs scheduledAt), recurring (needs
recurringConfig.frequency and time) or heartbeat (needs
recurringConfig.interval, optionally activeHoursStart/activeHoursEnd and
daysOfWeek). The task runs with the employee's chat settings unless
contextSettings is given.



# Delete a scheduled task.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/delete-a-scheduled-task

/openapi.json delete /api/v0/scheduled-tasks/{task_id}



# Duplicate a task's settings into a new task (no history).
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/duplicate-a-tasks-settings-into-a-new-task-no-history

/openapi.json post /api/v0/scheduled-tasks/{task_id}/duplicate
The copy starts paused so it can be renamed and reviewed before it begins
running.



# Ensure a template's persona exists as an employee on the company.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/ensure-a-templates-persona-exists-as-an-employee-on-the-company

/openapi.json post /api/v0/scheduled-task-templates/{template_id}/employee
Called when the user picks a template, so the create-task form opens with
the right AI employee already assigned. Reuses an existing company
employee with the persona's title; otherwise adds the persona to the
company carrying its name, photo, prompt and tags.



# Get a single scheduled task.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/get-a-single-scheduled-task

/openapi.json get /api/v0/scheduled-tasks/{task_id}



# Grade a run of this task and propose a more efficient prompt.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/grade-a-run-of-this-task-and-propose-a-more-efficient-prompt

/openapi.json post /api/v0/scheduled-tasks/{task_id}/evaluate
By default this triggers a fresh run. Pass executionId to evaluate a run
that already happened instead - these tasks take real actions, so re-running
just to get a score is often not what the user wants.

Pass notes to steer the rewrite ("stop it posting twice", "keep the tone").
The run shows what happened, not which parts the owner considers wrong.

A real run can take minutes, so this returns 202 immediately and the client
polls GET /scheduled-tasks/<task_id>/evaluate for the result.



# Latest evaluation for a task, including its status and any suggestion.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/latest-evaluation-for-a-task-including-its-status-and-any-suggestion

/openapi.json get /api/v0/scheduled-tasks/{task_id}/evaluate



# List execution history across every scheduled task in a company.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/list-execution-history-across-every-scheduled-task-in-a-company

/openapi.json get /api/v0/scheduled-task-executions



# List execution history for a specific task.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/list-execution-history-for-a-specific-task

/openapi.json get /api/v0/scheduled-tasks/{task_id}/executions



# List predefined task templates that prefill a new scheduled task.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/list-predefined-task-templates-that-prefill-a-new-scheduled-task

/openapi.json get /api/v0/scheduled-task-templates
Each template's persona (the AI employee that runs the task) is resolved
from the employee_personas table so the gallery can show the actual
person under the template title.



# List scheduled tasks for a company.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/list-scheduled-tasks-for-a-company

/openapi.json get /api/v0/scheduled-tasks



# List the daily logs (heartbeat status timeline) for a scheduled task.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/list-the-daily-logs-heartbeat-status-timeline-for-a-scheduled-task

/openapi.json get /api/v0/scheduled-tasks/{task_id}/daily-logs



# List visible connections for a template's required integration choices.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/list-visible-connections-for-a-templates-required-integration-choices

/openapi.json get /api/v0/scheduled-task-templates/{template_id}/setup



# Pause a scheduled task so no runs fire until it is resumed.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/pause-a-scheduled-task-so-no-runs-fire-until-it-is-resumed

/openapi.json post /api/v0/scheduled-tasks/{task_id}/pause



# Re-apply the task's source template at its latest version.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/re-apply-the-tasks-source-template-at-its-latest-version

/openapi.json post /api/v0/scheduled-tasks/{task_id}/upgrade-template
Overwrites the prompt and tool settings with the template's latest, keeping
the task's schedule, employee and name.



# Reset a task's history.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/reset-a-tasks-history

/openapi.json post /api/v0/scheduled-tasks/{task_id}/reset
Soft-deletes its executions, evaluations and daily logs/notes and zeroes
its run counters. The task's settings are untouched.



# Resume a paused scheduled task (recomputes its next run).
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/resume-a-paused-scheduled-task-recomputes-its-next-run

/openapi.json post /api/v0/scheduled-tasks/{task_id}/resume



# Run a scheduled task immediately.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/run-a-scheduled-task-immediately

/openapi.json post /api/v0/scheduled-tasks/{task_id}/run
Runs the same executor the scheduler uses (so it records an execution with
blocks/response the same way), in the background. A manual run executes even
if the task is paused and does NOT alter the task's normal schedule. Poll
the executions endpoint for the result.



# Update a scheduled task. Only the supplied fields change.
Source: https://docs.parallellabs.app/api-reference/scheduled-tasks/update-a-scheduled-task-only-the-supplied-fields-change

/openapi.json put /api/v0/scheduled-tasks/{task_id}



# Add a website to monitor and set its schedule + enabled modules.
Source: https://docs.parallellabs.app/api-reference/seo/add-a-website-to-monitor-and-set-its-schedule-+-enabled-modules

/openapi.json post /api/v0/seo/projects



# Call any DataForSEO v3 live endpoint with our credentials.
Source: https://docs.parallellabs.app/api-reference/seo/call-any-dataforseo-v3-live-endpoint-with-our-credentials

/openapi.json post /api/v0/seo/dataforseo
Endpoint reference: https://docs.dataforseo.com/v3/ - find the endpoint
there, then pass its path without the /v3/ prefix (e.g.
"backlinks/backlinks/live" for
https://docs.dataforseo.com/v3/backlinks/backlinks/live/) and the task
array exactly as that page documents it.
Only synchronous `live` endpoints are forwarded (plus on_page/instant_pages);
task-based, on_page and appendix endpoints are account-wide and rejected.
Billed at what the vendor actually charged for the call, converted to
credits. Failed or rejected tasks cost nothing and bill nothing.



# Compare authority across several domains.
Source: https://docs.parallellabs.app/api-reference/seo/compare-authority-across-several-domains

/openapi.json post /api/v0/seo/domain/compare
Intended for putting your own domain next to its competitors. Each domain
is billed separately when it is not already cached.



# Estimate the credits one run of a monitored website will cost.
Source: https://docs.parallellabs.app/api-reference/seo/estimate-the-credits-one-run-of-a-monitored-website-will-cost

/openapi.json post /api/v0/seo/projects/estimate
Priced from the vendor's list rates and the settings supplied (tracked
keywords, competitors, prompts, platforms, link and page caps). Actual
runs bill exact vendor spend, so treat this as a ceiling - modules that
read from cache first usually come in under it.



# Fetch one audit, including its issue summary and progress.
Source: https://docs.parallellabs.app/api-reference/seo/fetch-one-audit-including-its-issue-summary-and-progress

/openapi.json get /api/v0/seo/audits/{auditId}



# Find companies by the markup on their website.
Source: https://docs.parallellabs.app/api-reference/seo/find-companies-by-the-markup-on-their-website

/openapi.json post /api/v0/seo/domains-by-html-terms
Searches the technology index for domains whose live HTML contains the
given terms — a `<script src>` path, a vendor host, or any tag/attribute.
Because it matches the page's raw source, it finds every site carrying a
given widget or tech stack even when each site runs on its own domain, so
it is the way to build a list of companies using a specific tool.

Returns matching domains ranked by authority plus the total available for
paging via `offset`. Billed at the vendor's actual per-result cost.



# Flag keywords as saved.
Source: https://docs.parallellabs.app/api-reference/seo/flag-keywords-as-saved

/openapi.json post /api/v0/seo/keywords/saved
Sets the saved flag on keywords the company already has metrics for. Saving
the same set twice is a no-op rather than a duplicate.



# Get or create the project for a domain.
Source: https://docs.parallellabs.app/api-reference/seo/get-or-create-the-project-for-a-domain

/openapi.json post /api/v0/seo/projects/ensure
Module pages call this when a site is typed freeform: the domain gets a
blank project (manual schedule, no modules enabled) so the module can run
against it, without the user first setting up monitoring.



# Google Search Console connections installed for the company.
Source: https://docs.parallellabs.app/api-reference/seo/google-search-console-connections-installed-for-the-company

/openapi.json get /api/v0/seo/search-console/connections
Pass `domain` to flag which connection's site covers a project's domain.



# Latest anchor text, referring domains, and vendor backlink history.
Source: https://docs.parallellabs.app/api-reference/seo/latest-anchor-text-referring-domains-and-vendor-backlink-history

/openapi.json get /api/v0/seo/projects/{projectId}/link-profile



# Latest Lighthouse report (category scores, Core Web Vitals, failing
Source: https://docs.parallellabs.app/api-reference/seo/latest-lighthouse-report-category-scores-core-web-vitals-failing

/openapi.json get /api/v0/seo/projects/{projectId}/lighthouse
audits) plus per-scan scores over the last `days` for the trend chart.



# Latest organic competitor set and backlink gap for a project.
Source: https://docs.parallellabs.app/api-reference/seo/latest-organic-competitor-set-and-backlink-gap-for-a-project

/openapi.json get /api/v0/seo/projects/{projectId}/competitors



# List the company's keywords.
Source: https://docs.parallellabs.app/api-reference/seo/list-the-companys-keywords

/openapi.json get /api/v0/seo/keywords
Returns every keyword the company has metrics for - including the ones a
research run harvested - sorted by search volume. Pass savedOnly=true to
narrow to keywords a user explicitly kept.



# List the company's site audits, newest first.
Source: https://docs.parallellabs.app/api-reference/seo/list-the-companys-site-audits-newest-first

/openapi.json get /api/v0/seo/audits



# List the pages an audit crawled.
Source: https://docs.parallellabs.app/api-reference/seo/list-the-pages-an-audit-crawled

/openapi.json get /api/v0/seo/audits/{auditId}/pages
Pass issueId to see only the pages carrying one particular check, which is
how the summary drills into "which pages are missing a title".



# Look up authority and traffic for a domain.
Source: https://docs.parallellabs.app/api-reference/seo/look-up-authority-and-traffic-for-a-domain

/openapi.json get /api/v0/seo/domain/overview
Returns domain rating, estimated monthly organic traffic, total backlinks
and referring domain count. Served from cache when available.



# One Lighthouse scan in full, so an earlier snapshot can be inspected.
Source: https://docs.parallellabs.app/api-reference/seo/one-lighthouse-scan-in-full-so-an-earlier-snapshot-can-be-inspected

/openapi.json get /api/v0/seo/projects/{projectId}/lighthouse/{snapshotId}



# One project plus its latest per-module results.
Source: https://docs.parallellabs.app/api-reference/seo/one-project-plus-its-latest-per-module-results

/openapi.json get /api/v0/seo/projects/{projectId}



# Paginated AEO answers, filterable by platform/prompt.
Source: https://docs.parallellabs.app/api-reference/seo/paginated-aeo-answers-filterable-by-platformprompt

/openapi.json get /api/v0/seo/projects/{projectId}/ai-visibility
Each row is one AI platform's answer to one tracked prompt, already scored
for brand mention and citation.



# Paginated individual backlinks for a project's domain.
Source: https://docs.parallellabs.app/api-reference/seo/paginated-individual-backlinks-for-a-projects-domain

/openapi.json get /api/v0/seo/projects/{projectId}/backlinks



# Paginated list of a company's monitored SEO projects.
Source: https://docs.parallellabs.app/api-reference/seo/paginated-list-of-a-companys-monitored-seo-projects

/openapi.json get /api/v0/seo/projects



# Pause a project's schedule (no runs fire until resumed).
Source: https://docs.parallellabs.app/api-reference/seo/pause-a-projects-schedule-no-runs-fire-until-resumed

/openapi.json post /api/v0/seo/projects/{projectId}/pause



# Per-run AI visibility / citation / share-of-voice scores over time.
Source: https://docs.parallellabs.app/api-reference/seo/per-run-ai-visibility-citation-share-of-voice-scores-over-time

/openapi.json get /api/v0/seo/projects/{projectId}/ai-visibility/trend



# Per-run backlink totals + new/lost counts over time.
Source: https://docs.parallellabs.app/api-reference/seo/per-run-backlink-totals-+-newlost-counts-over-time

/openapi.json get /api/v0/seo/projects/{projectId}/backlink-history



# Queries a site already ranks for in Google Search Console, tagged for
Source: https://docs.parallellabs.app/api-reference/seo/queries-a-site-already-ranks-for-in-google-search-console-tagged-for

/openapi.json post /api/v0/seo/search-console/preview
seeding a monitored website's keywords and AEO prompts: which are questions
(prompts), branded, in striking distance, or already listed.

Body: {integrationId, domain, days?, trackedKeywords?, trackedPrompts?,
brandNames?} -- the lists are whatever the project form currently holds.
Free -- reads the company's Search Console connection.



# Rank snapshots over time, optionally filtered to one keyword.
Source: https://docs.parallellabs.app/api-reference/seo/rank-snapshots-over-time-optionally-filtered-to-one-keyword

/openapi.json get /api/v0/seo/projects/{projectId}/rank-history



# Remove a keyword from the saved list.
Source: https://docs.parallellabs.app/api-reference/seo/remove-a-keyword-from-the-saved-list

/openapi.json delete /api/v0/seo/keywords/{keywordId}/saved
Clears the saved flag only. The keyword and its metrics stay, so a later
run reuses them instead of re-buying the same term.



# Remove a monitored project.
Source: https://docs.parallellabs.app/api-reference/seo/remove-a-monitored-project

/openapi.json delete /api/v0/seo/projects/{projectId}



# Remove a site audit.
Source: https://docs.parallellabs.app/api-reference/seo/remove-a-site-audit

/openapi.json delete /api/v0/seo/audits/{auditId}



# Research search demand for a keyword.
Source: https://docs.parallellabs.app/api-reference/seo/research-search-demand-for-a-keyword

/openapi.json get /api/v0/seo/keywords/research
Expands the seed keyword into related terms and returns volume, CPC and
competition for each. Answers from cache when it can; only purchases the
terms that are missing, and reports whether anything was bought.



# Resume a paused project (recomputes its next run).
Source: https://docs.parallellabs.app/api-reference/seo/resume-a-paused-project-recomputes-its-next-run

/openapi.json post /api/v0/seo/projects/{projectId}/resume



# Run a project's modules now. Does not advance the schedule.
Source: https://docs.parallellabs.app/api-reference/seo/run-a-projects-modules-now-does-not-advance-the-schedule

/openapi.json post /api/v0/seo/projects/{projectId}/run
Body may carry `modules: [...]` (SEO_MODULE_* keys) to run only those -
enabled or not - so one report can be refreshed without re-running the
whole project. Without it, every enabled module runs.



# Run (or re-run) the AI readiness check for an audit.
Source: https://docs.parallellabs.app/api-reference/seo/run-or-re-run-the-ai-readiness-check-for-an-audit

/openapi.json post /api/v0/seo/audits/{auditId}/ai-readiness
Scores AI crawler access, llms.txt, Schema.org and response speed. Billed
separately from the crawl, and runs in the background.



# Start a site audit.
Source: https://docs.parallellabs.app/api-reference/seo/start-a-site-audit

/openapi.json post /api/v0/seo/audits
Crawls the domain from its homepage and checks each page for on-page SEO
problems. Returns as soon as the audit is queued - the crawl runs in the
background, so poll the audit until its status is complete.



# Stop a queued or running audit.
Source: https://docs.parallellabs.app/api-reference/seo/stop-a-queued-or-running-audit

/openapi.json post /api/v0/seo/audits/{auditId}/cancel
A queued audit stops immediately. A running one finishes its current wave
of pages and then stops, keeping whatever it already crawled - so the
partial result stays readable rather than being thrown away.



# Update a project's schedule, modules, tracked keywords, or caps.
Source: https://docs.parallellabs.app/api-reference/seo/update-a-projects-schedule-modules-tracked-keywords-or-caps

/openapi.json patch /api/v0/seo/projects/{projectId}



# Update the tags or note on a keyword.
Source: https://docs.parallellabs.app/api-reference/seo/update-the-tags-or-note-on-a-keyword

/openapi.json patch /api/v0/seo/keywords/{keywordId}



# Add a company-specific unsubscribe.
Source: https://docs.parallellabs.app/api-reference/sequences/add-a-company-specific-unsubscribe

/openapi.json post /api/v0/unsubscribes



# Add a member to a sequence.
Source: https://docs.parallellabs.app/api-reference/sequences/add-a-member-to-a-sequence

/openapi.json post /api/v0/sequences/{sequenceId}/members



# Add multiple company-specific unsubscribes from a CSV import.
Source: https://docs.parallellabs.app/api-reference/sequences/add-multiple-company-specific-unsubscribes-from-a-csv-import

/openapi.json post /api/v0/unsubscribes/bulk



# Append a single step to a sequence without resending the whole step list.
Source: https://docs.parallellabs.app/api-reference/sequences/append-a-single-step-to-a-sequence-without-resending-the-whole-step-list

/openapi.json post /api/v0/sequences/{sequenceId}/steps



# Approve executions across multiple sequences.
Source: https://docs.parallellabs.app/api-reference/sequences/approve-executions-across-multiple-sequences

/openapi.json post /api/v0/sequences/executions/approve



# Approve multiple executions for sending.
Source: https://docs.parallellabs.app/api-reference/sequences/approve-multiple-executions-for-sending

/openapi.json post /api/v0/sequences/{sequenceId}/executions/approve



# Bulk update multiple sender profiles with the same values.
Source: https://docs.parallellabs.app/api-reference/sequences/bulk-update-multiple-sender-profiles-with-the-same-values

/openapi.json post /api/v0/sequences/profiles/bulk-update



# Bulk update signatures for multiple sender profiles with variable replacement.
Source: https://docs.parallellabs.app/api-reference/sequences/bulk-update-signatures-for-multiple-sender-profiles-with-variable-replacement

/openapi.json post /api/v0/sequences/profiles/bulk-update-signature
Supported variables:
{{name}} - The sender profile's name
{{email}} - The sender profile's email address
{{first_name}} - First word of the name
{{last_name}} - Last word of the name (if multiple words)



# Check if an email is unsubscribed from a company.
Source: https://docs.parallellabs.app/api-reference/sequences/check-if-an-email-is-unsubscribed-from-a-company

/openapi.json post /api/v0/unsubscribes/check



# Count active members currently sitting on each of the given steps.
Source: https://docs.parallellabs.app/api-reference/sequences/count-active-members-currently-sitting-on-each-of-the-given-steps

/openapi.json get /api/v0/sequences/{sequenceId}/steps/member-counts
Powers the delete confirmation prompt, which has to tell the user how many
people are on a step before asking what should happen to them. Pass the step
ids as a repeated or comma-separated ``stepIds`` query param.



# Create a new sender profile.
Source: https://docs.parallellabs.app/api-reference/sequences/create-a-new-sender-profile

/openapi.json post /api/v0/sequences/profiles



# Create or update a sequence.
Source: https://docs.parallellabs.app/api-reference/sequences/create-or-update-a-sequence

/openapi.json post /api/v0/sequences
By default (``partial`` omitted or true) only the fields present in the body
are updated and every omitted field is left untouched -- this is the
recommended behavior for AI agents so a targeted edit never resets unrelated
settings. Send ``partial: false`` to do a full replace where omitted fields
fall back to their defaults (the frontend does this). Including ``steps``
replaces the full step list; to change a single step, prefer the
/sequences/<id>/steps endpoints. A ``steps`` array that drops every existing
step is rejected unless ``confirmDeleteAllSteps: true`` is also sent.



# Delete a sender profile.
Source: https://docs.parallellabs.app/api-reference/sequences/delete-a-sender-profile

/openapi.json delete /api/v0/sequences/profiles/{profileId}



# Delete a sequence by ID.
Source: https://docs.parallellabs.app/api-reference/sequences/delete-a-sequence-by-id

/openapi.json delete /api/v0/sequences/{sequenceId}



# Delete executions across multiple sequences.
Source: https://docs.parallellabs.app/api-reference/sequences/delete-executions-across-multiple-sequences

/openapi.json post /api/v0/sequences/executions/delete



# Delete multiple executions.
Source: https://docs.parallellabs.app/api-reference/sequences/delete-multiple-executions

/openapi.json post /api/v0/sequences/{sequenceId}/executions/delete



# Duplicate a sequence and its steps.
Source: https://docs.parallellabs.app/api-reference/sequences/duplicate-a-sequence-and-its-steps

/openapi.json post /api/v0/sequences/{sequenceId}/duplicate



# Edit an execution's subject or content.
Source: https://docs.parallellabs.app/api-reference/sequences/edit-an-executions-subject-or-content

/openapi.json post /api/v0/sequences/{sequenceId}/executions/{executionId}



# Execute approved executions across multiple sequences.
Source: https://docs.parallellabs.app/api-reference/sequences/execute-approved-executions-across-multiple-sequences

/openapi.json post /api/v0/sequences/executions/execute



# Execute (send) multiple approved executions.
Source: https://docs.parallellabs.app/api-reference/sequences/execute-send-multiple-approved-executions

/openapi.json post /api/v0/sequences/{sequenceId}/executions/execute



# Export all timeline events for a sequence to CSV.
Source: https://docs.parallellabs.app/api-reference/sequences/export-all-timeline-events-for-a-sequence-to-csv

/openapi.json get /api/v0/sequences/{sequenceId}/activity/export
Efficiently collects all activity data without pagination limits
and returns a downloadable CSV file.



# Generate a sequence using AI based on a goal.
Source: https://docs.parallellabs.app/api-reference/sequences/generate-a-sequence-using-ai-based-on-a-goal

/openapi.json post /api/v0/sequences/generate



# Get a single execution with its sequence and member details.
Source: https://docs.parallellabs.app/api-reference/sequences/get-a-single-execution-with-its-sequence-and-member-details

/openapi.json get /api/v0/sequences/executions/{executionId}



# Get a specific sequence by ID.
Source: https://docs.parallellabs.app/api-reference/sequences/get-a-specific-sequence-by-id

/openapi.json get /api/v0/sequences/{sequenceId}



# Get aggregate metrics for all sequences.
Source: https://docs.parallellabs.app/api-reference/sequences/get-aggregate-metrics-for-all-sequences

/openapi.json get /api/v0/sequences/metrics



# Get all available fields for variable insertion in a sequence.
Source: https://docs.parallellabs.app/api-reference/sequences/get-all-available-fields-for-variable-insertion-in-a-sequence

/openapi.json get /api/v0/sequences/{sequenceId}/available-fields
This includes standard member fields and custom userData fields from sequence members.

Returns:
- standardFields: List of standard member fields
- customFields: List of unique custom fields found in member userData



# Get all executions across all sequences for a company.
Source: https://docs.parallellabs.app/api-reference/sequences/get-all-executions-across-all-sequences-for-a-company

/openapi.json get /api/v0/sequences/executions/all



# Get comprehensive profile data for a sequence member.
Source: https://docs.parallellabs.app/api-reference/sequences/get-comprehensive-profile-data-for-a-sequence-member

/openapi.json get /api/v0/sequences/{sequenceId}/members/{memberId}/profile
Finds the member across different parts of the platform:
- Sequence member data
- Smart list appearances
- Agent conversations
- Other sequences



# Get hourly execution metrics for a specific sequence.
Source: https://docs.parallellabs.app/api-reference/sequences/get-hourly-execution-metrics-for-a-specific-sequence

/openapi.json get /api/v0/sequences/{sequenceId}/hourly-metrics



# Get hourly execution metrics for all sequences in a company.
Source: https://docs.parallellabs.app/api-reference/sequences/get-hourly-execution-metrics-for-all-sequences-in-a-company

/openapi.json get /api/v0/sequences/hourly-metrics



# Get list of unsubscribed emails for a company.
Source: https://docs.parallellabs.app/api-reference/sequences/get-list-of-unsubscribed-emails-for-a-company

/openapi.json get /api/v0/unsubscribes



# Get metrics broken down by step for a sequence.
Source: https://docs.parallellabs.app/api-reference/sequences/get-metrics-broken-down-by-step-for-a-sequence

/openapi.json get /api/v0/sequences/{sequenceId}/step-metrics



# Get paginated timeline events for a sequence across all members.
Source: https://docs.parallellabs.app/api-reference/sequences/get-paginated-timeline-events-for-a-sequence-across-all-members

/openapi.json get /api/v0/sequences/{sequenceId}/activity
Uses database-level pagination for efficiency with large datasets.
Includes detailed tracking events (opens, clicks) with IP, user agent, and clicked URLs.



# Get sequence step with sender profile and integration details.
Source: https://docs.parallellabs.app/api-reference/sequences/get-sequence-step-with-sender-profile-and-integration-details

/openapi.json get /api/v0/sequence-steps/{stepId}/details



# Get timeline of all actions for a sequence member.
Source: https://docs.parallellabs.app/api-reference/sequences/get-timeline-of-all-actions-for-a-sequence-member

/openapi.json get /api/v0/sequences/{sequenceId}/members/{sequenceMemberId}/timeline



# Import members from a smart list to a sequence with SSE progress.
Source: https://docs.parallellabs.app/api-reference/sequences/import-members-from-a-smart-list-to-a-sequence-with-sse-progress

/openapi.json post /api/v0/sequences/{sequenceId}/import-from-list



# List all executions for a sequence.
Source: https://docs.parallellabs.app/api-reference/sequences/list-all-executions-for-a-sequence

/openapi.json get /api/v0/sequences/{sequenceId}/executions



# List all members of a sequence.
Source: https://docs.parallellabs.app/api-reference/sequences/list-all-members-of-a-sequence

/openapi.json get /api/v0/sequences/{sequenceId}/members



# List all sender profiles for a company.
Source: https://docs.parallellabs.app/api-reference/sequences/list-all-sender-profiles-for-a-company

/openapi.json get /api/v0/sequences/profiles



# List all sequences for the current company.
Source: https://docs.parallellabs.app/api-reference/sequences/list-all-sequences-for-the-current-company

/openapi.json get /api/v0/sequences



# List the available sequence step (action) types and how to configure each.
Source: https://docs.parallellabs.app/api-reference/sequences/list-the-available-sequence-step-action-types-and-how-to-configure-each

/openapi.json get /api/v0/sequences/step-types
This is the source of truth for what steps a sequence can contain. Each entry
documents the actionType to use, what it does, which integration it needs (and
how to find valid IDs), and the actionParameters fields it accepts — read it
before building or editing steps.



# Manually record a metric (booking or reply) for a member at a specific step.
Source: https://docs.parallellabs.app/api-reference/sequences/manually-record-a-metric-booking-or-reply-for-a-member-at-a-specific-step

/openapi.json post /api/v0/sequences/{sequenceId}/manual-metric
Used when a recipient takes a tracked action off-channel and the user wants to
credit it to the correct sequence step for attribution.



# Partially update a single step. Only the fields present in the body are
Source: https://docs.parallellabs.app/api-reference/sequences/partially-update-a-single-step-only-the-fields-present-in-the-body-are

/openapi.json patch /api/v0/sequences/{sequenceId}/steps/{stepId}
changed; omitted fields keep their saved value, and the other steps in the
sequence are left untouched.



# Preview an execution's rendered content.
Source: https://docs.parallellabs.app/api-reference/sequences/preview-an-executions-rendered-content

/openapi.json get /api/v0/sequences/{sequenceId}/executions/{executionId}/preview



# Rebuild an execution's content using AI personalization.
Source: https://docs.parallellabs.app/api-reference/sequences/rebuild-an-executions-content-using-ai-personalization

/openapi.json post /api/v0/sequences/executions/{executionId}/rebuild
Regenerates the subject and content for the execution based on the step's settings.



# Remove a member from a sequence.
Source: https://docs.parallellabs.app/api-reference/sequences/remove-a-member-from-a-sequence

/openapi.json post /api/v0/sequences/{sequenceId}/members/{sequenceMemberId}/remove



# Remove an email from the unsubscribe list (resubscribe).
Source: https://docs.parallellabs.app/api-reference/sequences/remove-an-email-from-the-unsubscribe-list-resubscribe

/openapi.json delete /api/v0/unsubscribes/{unsubscribeId}



# Rerun an execution - reset its status and optionally regenerate content.
Source: https://docs.parallellabs.app/api-reference/sequences/rerun-an-execution--reset-its-status-and-optionally-regenerate-content

/openapi.json post /api/v0/sequences/executions/{executionId}/rerun
This allows failed or executed steps to be retried.



# Send a test email to a specified address with pre-generated content.
Source: https://docs.parallellabs.app/api-reference/sequences/send-a-test-email-to-a-specified-address-with-pre-generated-content

/openapi.json post /api/v0/sequences/{sequenceId}/steps/{stepId}/test/send



# Skip executions (typically validation failures).
Source: https://docs.parallellabs.app/api-reference/sequences/skip-executions-typically-validation-failures

/openapi.json post /api/v0/sequences/executions/skip
Marks them as executed so the sequence can continue,
but also marks them as failed with the skip reason.



# Soft-delete a single step and renumber the remaining steps.
Source: https://docs.parallellabs.app/api-reference/sequences/soft-delete-a-single-step-and-renumber-the-remaining-steps

/openapi.json delete /api/v0/sequences/{sequenceId}/steps/{stepId}
Accepts an optional ``memberDisposition`` (with ``targetStepId`` for
'specific_step') deciding what happens to active members sitting on the
step. Defaults to moving them to the next surviving step.



# Test a sequence step with a specific member to preview AI-generated content.
Source: https://docs.parallellabs.app/api-reference/sequences/test-a-sequence-step-with-a-specific-member-to-preview-ai-generated-content

/openapi.json post /api/v0/sequences/{sequenceId}/steps/{stepId}/test



# Trigger the next step for a sequence member.
Source: https://docs.parallellabs.app/api-reference/sequences/trigger-the-next-step-for-a-sequence-member

/openapi.json post /api/v0/sequences/{sequenceId}/members/{sequenceMemberId}/trigger



# Update a sender profile.
Source: https://docs.parallellabs.app/api-reference/sequences/update-a-sender-profile

/openapi.json put /api/v0/sequences/profiles/{profileId}



# List available agent skills
Source: https://docs.parallellabs.app/api-reference/skills/list-available-agent-skills

/openapi.json get /api/v0/skills
Returns each skill's slug and title. Fetch a skill's full instructions with
getSkill. Capabilities that have a related skill name its slug in their
listing (relatedSkills).



# Read a skill's full instructions
Source: https://docs.parallellabs.app/api-reference/skills/read-a-skills-full-instructions

/openapi.json get /api/v0/skills/{slug}
Returns the skill's Markdown instruction text. Read the relevant skill
before using a capability's operations — e.g. read "website-editing" before
editing a website, so you edit site.json instead of hand-writing tags the
platform generates automatically.



# Create a new support ticket.
Source: https://docs.parallellabs.app/api-reference/support/create-a-new-support-ticket

/openapi.json post /api/v0/support/tickets
Submit a bug report, feature request, question, or other support issue.
The ticket is scoped to the authenticated user and company.



# Get a single support ticket by ID.
Source: https://docs.parallellabs.app/api-reference/support/get-a-single-support-ticket-by-id

/openapi.json get /api/v0/support/tickets/{ticketId}



# List all messages on a support ticket, ordered oldest to newest.
Source: https://docs.parallellabs.app/api-reference/support/list-all-messages-on-a-support-ticket-ordered-oldest-to-newest

/openapi.json get /api/v0/support/tickets/{ticketId}/messages



# List support tickets accessible to the authenticated user.
Source: https://docs.parallellabs.app/api-reference/support/list-support-tickets-accessible-to-the-authenticated-user

/openapi.json get /api/v0/support/tickets
Regular users only see their own tickets. Supports filtering by status,
ticket type, severity, free-text query, and pagination.



# Post a new message (reply) to an existing support ticket.
Source: https://docs.parallellabs.app/api-reference/support/post-a-new-message-reply-to-an-existing-support-ticket

/openapi.json post /api/v0/support/tickets/{ticketId}/messages



# Update fields on an existing support ticket.
Source: https://docs.parallellabs.app/api-reference/support/update-fields-on-an-existing-support-ticket

/openapi.json put /api/v0/support/tickets/{ticketId}
Include only the fields you want to change. Status and response fields
may be restricted to support/admin users.



# Create a new tag.
Source: https://docs.parallellabs.app/api-reference/tags/create-a-new-tag

/openapi.json post /api/v0/tags



# Delete a tag.
Source: https://docs.parallellabs.app/api-reference/tags/delete-a-tag

/openapi.json delete /api/v0/tags/{tag_id}



# Get all tags for a company.
Source: https://docs.parallellabs.app/api-reference/tags/get-all-tags-for-a-company

/openapi.json get /api/v0/tags



# Update a tag.
Source: https://docs.parallellabs.app/api-reference/tags/update-a-tag

/openapi.json put /api/v0/tags/{tag_id}



# Daily credit usage across an organization for the last N days (default 30, max 90), stacked by user.
Source: https://docs.parallellabs.app/api-reference/usage/daily-credit-usage-across-an-organization-for-the-last-n-days-default-30-max-90-stacked-by-user

/openapi.json get /api/v0/organizations/{orgId}/usage/histogram



# Get aggregate usage totals across all companies in an organization.
Source: https://docs.parallellabs.app/api-reference/usage/get-aggregate-usage-totals-across-all-companies-in-an-organization

/openapi.json get /api/v0/organizations/{orgId}/usage/history/summary



# Get aggregate usage totals for the authenticated user.
Source: https://docs.parallellabs.app/api-reference/usage/get-aggregate-usage-totals-for-the-authenticated-user

/openapi.json get /api/v0/usage/history/summary
Returns total credits/tokens/requests plus breakdowns by model and feature.
Filtered to the user; company API keys add a company filter.



# Get daily usage histogram for the last 30 days.
Source: https://docs.parallellabs.app/api-reference/usage/get-daily-usage-histogram-for-the-last-30-days

/openapi.json get /api/v0/usage/histogram/last_30_days



# Get estimated hours saved by automation for a company.
Source: https://docs.parallellabs.app/api-reference/usage/get-estimated-hours-saved-by-automation-for-a-company

/openapi.json get /api/v0/usage/hours-saved
Calculates time saved based on automated actions performed in the last 30 days,
using industry-standard time estimates for each action type.



# Get organization usage bucketed over a time range, for the chart above the log.
Source: https://docs.parallellabs.app/api-reference/usage/get-organization-usage-bucketed-over-a-time-range-for-the-chart-above-the-log

/openapi.json get /api/v0/organizations/{orgId}/usage/activity
Takes the same `startDate`, `endDate`, `userUid` and `feature` filters as
the usage history endpoint, so the chart always describes the same rows the
log is showing. The bucket width adapts to the range.



# Get paginated usage history rows across all companies in an organization.
Source: https://docs.parallellabs.app/api-reference/usage/get-paginated-usage-history-rows-across-all-companies-in-an-organization

/openapi.json get /api/v0/organizations/{orgId}/usage/history
Caller must be an organization owner or admin. Optional `userUid` and
`feature` query parameters narrow results within the org.



# Get paginated usage history rows for the authenticated user.
Source: https://docs.parallellabs.app/api-reference/usage/get-paginated-usage-history-rows-for-the-authenticated-user

/openapi.json get /api/v0/usage/history
Filtered to the user's own usage. When authenticated via a company API key,
results are additionally restricted to that company.



# Get top 10 users by credit usage in a company.
Source: https://docs.parallellabs.app/api-reference/usage/get-top-10-users-by-credit-usage-in-a-company

/openapi.json get /api/v0/usage/leaderboard



# Get usage breakdown by feature for the last 30 days.
Source: https://docs.parallellabs.app/api-reference/usage/get-usage-breakdown-by-feature-for-the-last-30-days

/openapi.json get /api/v0/usage/features/last_30_days



# Get usage information including credits, features, and quotas.
Source: https://docs.parallellabs.app/api-reference/usage/get-usage-information-including-credits-features-and-quotas

/openapi.json get /api/v0/usage



# Create a video using AI generation.
Source: https://docs.parallellabs.app/api-reference/video/create-a-video-using-ai-generation

/openapi.json post /api/v0/videos
Queues a video generation job for the provided prompt and fal.ai model, and
returns immediately with the video id and a 'generating' status. Generation
takes minutes, so poll the status endpoint to find out when it is ready.



# Delete a video by ID.
Source: https://docs.parallellabs.app/api-reference/video/delete-a-video-by-id

/openapi.json delete /api/v0/videos/{videoId}



# Get available video models.
Source: https://docs.parallellabs.app/api-reference/video/get-available-video-models

/openapi.json get /api/v0/videos/models
Returns a list of available video generation models from fal.ai.
This endpoint is public and does not require authentication.



# Get available video styles (deprecated).
Source: https://docs.parallellabs.app/api-reference/video/get-available-video-styles-deprecated

/openapi.json get /api/v0/videos/styles
This endpoint is deprecated as fal.ai models handle styles internally.
Returns an empty object for backwards compatibility.



# Get video by ID.
Source: https://docs.parallellabs.app/api-reference/video/get-video-by-id

/openapi.json get /api/v0/videos/{videoId}
Returns the video data including URL and status.
This endpoint is public and does not require authentication.



# Get video generation status.
Source: https://docs.parallellabs.app/api-reference/video/get-video-generation-status

/openapi.json get /api/v0/videos/{videoId}/status
Returns the current status of video generation. When complete,
includes the video URL.



# List all videos.
Source: https://docs.parallellabs.app/api-reference/video/list-all-videos

/openapi.json get /api/v0/videos
Returns a paginated list of videos for the company.



# Create a custom voice using ElevenLabs Voice Cloning (IVC) API
Source: https://docs.parallellabs.app/api-reference/voices/create-a-custom-voice-using-elevenlabs-voice-cloning-ivc-api

/openapi.json post /api/v0/voices
Request body (multipart/form-data):
- companyId: Company ID (required)
- name: Voice name (required)
- description: Voice description (optional)
- files: Audio files for voice cloning (required, one or more)
- remove_background_noise: Whether to remove background noise (optional, default: true)

ElevenLabs API Reference:
https://elevenlabs.io/docs/api-reference/voices/ivc/create



# Delete a custom voice (soft delete)
Source: https://docs.parallellabs.app/api-reference/voices/delete-a-custom-voice-soft-delete

/openapi.json delete /api/v0/voices/{voice_id}



# Fetch the full ElevenLabs voice library
Source: https://docs.parallellabs.app/api-reference/voices/fetch-the-full-elevenlabs-voice-library

/openapi.json get /api/v0/voices/elevenlabs/library
This endpoint fetches voices from ElevenLabs API and caches them



# Get a specific voice by ID
Source: https://docs.parallellabs.app/api-reference/voices/get-a-specific-voice-by-id

/openapi.json get /api/v0/voices/{voice_id}



# List all voices available to the user (company-specific by default)
Source: https://docs.parallellabs.app/api-reference/voices/list-all-voices-available-to-the-user-company-specific-by-default

/openapi.json get /api/v0/voices
Query params:
- companyId: Company ID (optional if includePublic=true)
- page: Page number (default: 1)
- pageSize: Page size (default: 100)
- includePublic: Whether to include public voices (default: False)



# Test a voice by generating audio from text
Source: https://docs.parallellabs.app/api-reference/voices/test-a-voice-by-generating-audio-from-text

/openapi.json post /api/v0/voices/{voice_id}/test
Request body:
- companyId: Company ID (required)
- text: Text to convert to speech (required)
- ttsModel: ElevenLabs model to synthesize with (optional)

Returns audio/mpeg binary data.



# Update a custom voice
Source: https://docs.parallellabs.app/api-reference/voices/update-a-custom-voice

/openapi.json put /api/v0/voices/{voice_id}



# Capture a screenshot of a web page
Source: https://docs.parallellabs.app/api-reference/web/capture-a-screenshot-of-a-web-page

/openapi.json post /api/v0/web/screenshot
Renders the page, captures it as a PNG (full scrollable page by default,
viewport only with fullPage=false), saves it to the image library, and
returns a public image URL. Costs 1 data credit per call.



# Get full details for a place
Source: https://docs.parallellabs.app/api-reference/web/get-full-details-for-a-place

/openapi.json get /api/v0/web/maps/place
Returns name, address, phone, website, hours, rating, and coordinates
for one place. Provide placeId or dataId from a places search result.
Costs 1 data credit per call.



# Get photos for a place
Source: https://docs.parallellabs.app/api-reference/web/get-photos-for-a-place

/openapi.json get /api/v0/web/maps/photos
Returns photo thumbnails and full-size image URLs for a place identified
by the dataId from a places search result. Filter by categoryId and
paginate with nextPageToken. Costs 1 data credit per call.



# Get reviews for a place
Source: https://docs.parallellabs.app/api-reference/web/get-reviews-for-a-place

/openapi.json get /api/v0/web/maps/reviews
Returns reviews (rating, text, date, reviewer) for a place identified by
the dataId from a places search result. Sort with sortBy and paginate
with nextPageToken. Costs 1 data credit per call.



# Map a phrase to trending search topics
Source: https://docs.parallellabs.app/api-reference/web/map-a-phrase-to-trending-search-topics

/openapi.json get /api/v0/web/trends/autocomplete
Disambiguates a phrase into the trending topics it could refer to (e.g.
'jaguar' the animal vs the car brand), each with a topic title and type.
Costs 1 data credit per call.



# Read a web page
Source: https://docs.parallellabs.app/api-reference/web/read-a-web-page

/openapi.json get /api/v0/web/scrape
Fetches the page and returns it as clean markdown (default) or raw HTML
via responseType. JavaScript-rendered pages and hard-to-scrape sites,
including LinkedIn, are handled automatically. Costs 1 data credit per
call.



# Search job listings
Source: https://docs.parallellabs.app/api-reference/web/search-job-listings

/openapi.json get /api/v0/web/jobs
Returns job listings (title, company, location, description) for a query
like 'software engineer in New York'. Paginate with nextPageToken from a
previous response. Costs 1 data credit per call.



# Search local businesses and places
Source: https://docs.parallellabs.app/api-reference/web/search-local-businesses-and-places

/openapi.json get /api/v0/web/maps/search
Finds businesses matching a query like 'coffee shops in Austin', with
address, rating, phone, and coordinates. Each result includes the
place_id and data_id used by the place-details, reviews, and photos
endpoints. Paginate with page (0, 20, 40, ...). Costs 1 data credit per
call.



# Search recent news
Source: https://docs.parallellabs.app/api-reference/web/search-recent-news

/openapi.json get /api/v0/web/news
Returns titles, URLs, publishers, and dates for matching news coverage.
Narrow the window with freshness (pd = past day, pw = past week, pm =
past month, py = past year). Costs 1 data credit per call.



# Search the web
Source: https://docs.parallellabs.app/api-reference/web/search-the-web

/openapi.json get /api/v0/web/search
Returns titles, URLs, snippets, and dates for the best-matching pages.
Pass a result URL to the scrape endpoint to read the page in full.
Costs 1 data credit per call.



# View a web page or an image and answer a question about it.
Source: https://docs.parallellabs.app/api-reference/web/view-a-web-page-or-an-image-and-answer-a-question-about-it

/openapi.json post /api/v0/web/view
Pass a page `url` (screenshotted in a real browser) OR a direct `imageUrl`
(viewed as-is), plus a `query`. A vision model answers in text based only on
what it can see. Costs 1 data credit per successful call. This is the same
unified capability the AI-employee `take_website_screenshot` / `view_image`
tools use (app/utils/visual_qa.py).



# Audit a website's SEO and AEO
Source: https://docs.parallellabs.app/api-reference/websites/audit-a-websites-seo-and-aeo

/openapi.json get /api/v0/websites/{websiteId}/seo-audit
Returns a score plus per-page state and issues (worst first): titles, meta
descriptions, canonical/robots, alt text, heading hierarchy, structured data,
Open Graph, broken internal links, orphan pages, and FAQ-schema opportunities.
Read-only — pass no branch to audit the live site.



# Delete a website file on a review branch
Source: https://docs.parallellabs.app/api-reference/websites/delete-a-website-file-on-a-review-branch

/openapi.json delete /api/v0/websites/{websiteId}/file
Removes a file on a review branch (a branch tombstone). The branch is
REQUIRED and must not be "main".



# Diff a review branch against the live site
Source: https://docs.parallellabs.app/api-reference/websites/diff-a-review-branch-against-the-live-site

/openapi.json get /api/v0/websites/{websiteId}/branches/{branch}/diff
Returns the changed files and their before/after, i.e. exactly what the
owner will review before approving.



# Lighthouse-score a published page
Source: https://docs.parallellabs.app/api-reference/websites/lighthouse-score-a-published-page

/openapi.json get /api/v0/websites/{websiteId}/lighthouse
Returns performance, accessibility, best-practices and SEO scores plus Core
Web Vitals for one live page. Only works after the site is published; for a
pre-publish check use the seo-audit endpoint instead.



# List a website's pages
Source: https://docs.parallellabs.app/api-reference/websites/list-a-websites-pages

/openapi.json get /api/v0/websites/{websiteId}/pages
Each entry maps a public route (e.g. /about) to the file that renders it,
so you know which files to read and where fixes belong.



# List a website's review branches
Source: https://docs.parallellabs.app/api-reference/websites/list-a-websites-review-branches

/openapi.json get /api/v0/websites/{websiteId}/branches
Shows the review branches you (or the owner) have created and how many files
each changes, so you can see what is already staged before you audit again.



# List the company's websites
Source: https://docs.parallellabs.app/api-reference/websites/list-the-companys-websites

/openapi.json get /api/v0/websites
Start here: returns each website's id and name so you can audit or optimize
one. branchCount is how many review branches already have unmerged changes.



# Read one website file
Source: https://docs.parallellabs.app/api-reference/websites/read-one-website-file

/openapi.json get /api/v0/websites/{websiteId}/file
Returns the full content of a file so you can inspect it before proposing a
change. Pass a branch to read your in-progress version, or omit it for the
live file.



# Validate a website renders cleanly
Source: https://docs.parallellabs.app/api-reference/websites/validate-a-website-renders-cleanly

/openapi.json get /api/v0/websites/{websiteId}/validate
Reports broken includes, missing layouts, template cycles and invalid
site.json. Run this after writing fixes to a branch to confirm nothing broke.



# Write a website file onto a review branch
Source: https://docs.parallellabs.app/api-reference/websites/write-a-website-file-onto-a-review-branch

/openapi.json put /api/v0/websites/{websiteId}/file
Applies a fix. The branch is REQUIRED and must not be "main": the change
lands on a review branch the owner approves, never the live site. The branch
is created automatically on the first write. The site cannot be published
from the API — draft-only.



# Apps
Source: https://docs.parallellabs.app/apps

Open and use the apps you have published and installed.

Your home for the apps you've published and installed. Open any app right inside the platform — no code, no build step, just click and use.

## What Is the Apps Section?

The **Apps** section in the sidebar lists the apps available to you and your team:

* **Apps you've published** in [App Builder](/inbound/app-builder)
* **Apps added for your company** — including linked external tools

It's the "use" side of apps. To *build or edit* an app, use [App Builder](/inbound/app-builder); to *run* one, open it here.

## Opening an App

1. Select your company from the sidebar
2. Expand the **Apps** section
3. Click an app to open it in the main area

Each app shows its own icon and name. Some apps open in place; others may open in a new tab, depending on how they're set up.

## Apps vs. App Builder

|                  | App Builder                               | Apps                     |
| ---------------- | ----------------------------------------- | ------------------------ |
| **Purpose**      | Create and edit apps                      | Open and use apps        |
| **What you see** | A full editor (chat, files, live preview) | The running app          |
| **Who uses it**  | You (and company members, if shared)      | Anyone with access       |
| **Build step**   | You click **Publish**                     | None — just open and use |

## Managing Which Apps Appear

Published apps show up here automatically based on their sharing settings and whether they're enabled for your company. To change an app itself — rename it, change its icon, adjust sharing, or publish a new version — open it in [App Builder](/inbound/app-builder).

## Next Steps

* [Build an App](/inbound/app-builder)
* [Create a Landing Page](/inbound/landing-pages)


# Parallel AI REST API Authentication and Access Keys
Source: https://docs.parallellabs.app/authentication

Parallel AI supports Company API Keys and Personal Access Keys. Learn how to generate each key type and pass it correctly in your API requests.

## Two Key Types

Parallel AI offers two types of API keys to fit different use cases. Understanding which to use will keep your integrations secure and maintainable.

| Use Case                    | Recommended Key Type |
| --------------------------- | -------------------- |
| Single company integration  | Company API Key      |
| Multi-company dashboard     | Personal Access Key  |
| CI/CD pipelines             | Company API Key      |
| Personal automation scripts | Personal Access Key  |
| Third-party app integration | Company API Key      |

### Company API Key

A Company API Key is **scoped to a single company workspace**. When you use this key, all API calls automatically operate within that company's context — no extra headers needed.

Best for:

* Server-side integrations tied to one company
* CI/CD pipelines and deployment automation
* Third-party apps that connect to a single workspace

### Personal Access Key (`pak_`)

A Personal Access Key is tied to **your user account** and works across every company you belong to. All Personal Access Keys begin with the prefix `pak_`. Because they aren't scoped to a single company, you must specify which company you want to act on by passing a `X-Company-ID` header (or `companyId` query parameter) with each request.

Best for:

* Multi-company dashboards and admin tooling
* Personal automation scripts
* Local development where you switch between workspaces

<Note>
  Personal Access Keys inherit your user-level permissions. If your account is removed from a company, your Personal Access Key will immediately lose access to that company's resources.
</Note>

## Getting a Company API Key

<Steps>
  <Step title="Log in to your dashboard">
    Visit [https://web.parallellabs.app](https://web.parallellabs.app) and sign in to your account.
  </Step>

  <Step title="Navigate to the Integrations page">
    From the left sidebar, click **Integrations**.
  </Step>

  <Step title="Generate a new key">
    Find the **API Keys** section and click **Generate New Key**. Give it a descriptive name so you can identify it later.
  </Step>

  <Step title="Copy and store your key">
    Copy the key shown in the dialog. **This is the only time it will be displayed** — store it securely in a password manager, secrets vault, or environment variable before closing.
  </Step>
</Steps>

### Using a Company API Key

You can pass your Company API Key in one of two ways:

<Tabs>
  <Tab title="X-API-Key header">
    ```bash theme={null}
    curl -X GET "https://api.parallellabs.app/api/v0/agents" \
      -H "X-API-Key: YOUR_COMPANY_API_KEY"
    ```
  </Tab>

  <Tab title="Authorization header (OpenAI-compatible)">
    If you're using an OpenAI-compatible client that expects a Bearer token, you can pass your Company API Key as the bearer value:

    ```bash theme={null}
    curl -X GET "https://api.parallellabs.app/api/v0/agents" \
      -H "Authorization: Bearer YOUR_COMPANY_API_KEY"
    ```

    This is especially useful when configuring the `openai` Python or Node.js library, which sets `Authorization: Bearer <api_key>` automatically.
  </Tab>
</Tabs>

## Getting a Personal Access Key

<Steps>
  <Step title="Log in to your dashboard">
    Visit [https://web.parallellabs.app](https://web.parallellabs.app) and sign in to your account.
  </Step>

  <Step title="Open your profile">
    Click your **avatar or profile icon** in the top-right corner and select **Profile**.
  </Step>

  <Step title="Find Personal Access Keys">
    Scroll to the **Personal Access Keys** section on your profile page.
  </Step>

  <Step title="Create a new key">
    Click **Create New Key**. Give it a descriptive name (e.g. `local-dev` or `multi-company-script`). Optionally, set an **expiration date** to automatically revoke the key after a set period.
  </Step>

  <Step title="Copy and store your key">
    Your new key — beginning with `pak_` — will be displayed once. Copy it now and store it securely. **It won't be shown again.**
  </Step>
</Steps>

### Using a Personal Access Key

Personal Access Keys must be accompanied by a company identifier so the API knows which workspace to act on. You can pass your PAK using either the `Authorization` header or the `X-API-Key` header, and supply the company context via the `X-Company-ID` header or `companyId` query parameter.

**Option 1 — `Authorization` header with `X-Company-ID` (recommended):**

```bash theme={null}
curl -X GET "https://api.parallellabs.app/api/v0/agents" \
  -H "Authorization: Bearer pak_your_personal_access_key" \
  -H "X-Company-ID: your_company_id"
```

**Option 2 — `X-API-Key` header with `X-Company-ID`:**

```bash theme={null}
curl -X GET "https://api.parallellabs.app/api/v0/agents" \
  -H "X-API-Key: pak_your_personal_access_key" \
  -H "X-Company-ID: your_company_id"
```

**Option 3 — `companyId` query parameter:**

```bash theme={null}
curl -X GET "https://api.parallellabs.app/api/v0/agents?companyId=your_company_id" \
  -H "Authorization: Bearer pak_your_personal_access_key"
```

<Tip>
  The `X-Company-ID` header approach is preferred for production use — it keeps your company ID out of server logs and browser history that may record request URLs.
</Tip>

## Finding Your Company ID

You need your Company ID when using a Personal Access Key. You can find it in two places:

* **Dashboard URL** — when you're viewing a company in the dashboard, the company ID appears in the URL (e.g. `https://web.parallellabs.app/companies/YOUR_COMPANY_ID/...`).
* **Settings page** — navigate to your company's **Settings** page; the Company ID is listed there and can be copied directly.

## SSO Login

Enterprise customers can authenticate through their organization's identity provider using Single Sign-On (SSO).

Navigate to [https://web.parallellabs.app/login/sso](https://web.parallellabs.app/login/sso) and enter your **company domain** to be redirected to your SSO provider. Contact your Parallel AI account manager to enable SSO for your organization.

## Security Best Practices

<Warning>
  API keys grant access to your company's data and resources. Treat them like passwords.
</Warning>

* **Never commit API keys to source control.** Use environment variables or a secrets manager instead. Add `.env` to your `.gitignore`.

* **Rotate keys immediately if exposed.** If a key is accidentally published or shared, revoke it in the dashboard right away and generate a replacement.

* **Use environment variables in your code:**

  ```bash theme={null}
  export PARALLEL_API_KEY="YOUR_API_KEY"
  ```

  ```python theme={null}
  import os
  api_key = os.environ["PARALLEL_API_KEY"]
  ```

* **Give keys descriptive names.** Names like `production-crm-integration` or `ci-deploy-pipeline` make it easy to identify the purpose of each key and safely revoke the right one if needed.

* **Set expiration dates on Personal Access Keys** when you only need them for a limited time — for example, a temporary migration script.

* **Apply least privilege** — use Company API Keys for integrations scoped to a single workspace rather than Personal Access Keys with broader access, wherever possible.


# Content Engine
Source: https://docs.parallellabs.app/engage/content-engine

Plan what to say, generate it on brand, and produce the images, audio, and video that go with it.

The Content Engine is the largest section in the platform. It covers everything from deciding who you're writing for through to publishing a finished post with its artwork.

Its menu is organized in four groups, and you don't have to use them in order — start wherever it's useful.

## Brand

Set up once, then everything you generate draws on it.

* **[Ideal Customer Profile](/engage/content-engine/icp)** — Who you're writing for.
* **[Content Strategy](/engage/content-engine/content-strategy)** — What topics you'll cover, mapped across the funnel.
* **[Brand Voice](/engage/content-engine/brand-voice)** — How your writing should sound.
* **[Brand Images](/engage/content-engine/brand-images)** — Reusable styles and frames so visuals stay on-brand.

All four are optional. Content generated without them still works — it just won't be as specific to your business.

## Content

* **[Campaigns](/engage/content-engine/content-campaigns)** — Recurring content programs that generate posts on a schedule.
* **[Content](/engage/content-engine/content)** — Every piece you've created, in a list or a calendar, with approvals and publishing.

## SEO & AEO

Research what people actually search for — and what AI assistants actually cite.

* **[Keyword Research](/engage/content-engine/seo-keyword-research)** — Search demand, cost per click, and competition for any keyword.
* **[Domain Overview](/engage/content-engine/seo-domain-overview)** — Authority, traffic, and backlinks for your domain and your competitors'.
* **[Keywords](/engage/content-engine/seo-keywords)** — Every keyword you've looked up, with the ones you've starred.
* **[Site Audit](/engage/content-engine/seo-site-audit)** — Crawl your site and find on-page SEO problems.
* **[Opportunities](/engage/content-engine/seo-opportunities)** — Ranked content opportunities with real search volumes and ready-to-write briefs.

## Media

* **[Images](/engage/content-engine/images)** — Generate and edit images from text prompts.
* **[Canvas](/engage/content-engine/canvas)** — Design social graphics, story art, and link previews with an AI design assistant.
* **[Audio](/engage/content-engine/audio)** — Turn text into natural-sounding voiceovers.
* **[Video](/engage/content-engine/video)** — Generate short videos from text prompts.
* **[Voices](/engage/content-engine/voices)** — Clone custom voices or use ready-made ones.

## A Suggested Order

If you're starting from nothing, this order builds on itself:

1. **[Ideal Customer Profile](/engage/content-engine/icp)** — who are you talking to?
2. **[Content Strategy](/engage/content-engine/content-strategy)** — what themes will you own?
3. **[Brand Voice](/engage/content-engine/brand-voice)** — how should it sound?
4. **[Brand Images](/engage/content-engine/brand-images)** — what should it look like?
5. **[Campaigns](/engage/content-engine/content-campaigns)** — how often does it go out?
6. **[Content](/engage/content-engine/content)** — write, review, publish.

Steps 1–4 are optional groundwork. If you just want a blog post right now, go straight to [Content](/engage/content-engine/content) and use **Create from Idea**.

## Where SEO Fits

The SEO & AEO group answers *what should we write about* rather than *how do we write it*. The usual path is [Opportunities](/engage/content-engine/seo-opportunities) to find a topic worth covering, then [Content](/engage/content-engine/content) to write it — the opportunity comes with a brief you can generate straight from.


# Audio
Source: https://docs.parallellabs.app/engage/content-engine/audio

Turn text into natural-sounding voiceovers and audio.

Turn text into natural-sounding speech using AI voice technology. Create voiceovers for videos, podcasts, training materials, phone systems, and more.

## What Can You Create?

* Video narration and voiceovers
* Podcast intros and outros
* Training and educational content
* IVR phone messages
* Audiobook content
* Voice announcements
* Accessibility audio versions of content

## Creating Your First Audio

### Step 1: Open Audio Generation

1. Select your company from the sidebar
2. Click **Audio** (under the Content Engine menu)
3. Click **Generate Audio** (or the **+** button)

### Step 2: Enter Your Text

Type or paste the text you want converted to speech:

```
Welcome to our company! We're excited to have you here.
In this video, we'll walk you through the key features
that will help you get started quickly.
```

### Step 3: Choose a Voice

Browse available voices and pick one that fits your needs:

* Male or female
* Different accents (American, British, etc.)
* Various personality types (friendly, professional, energetic)

You can choose from ready-made voices or select a **custom voice you've cloned** in the [Voices](/engage/content-engine/voices) section — your cloned voices appear right alongside the ready-made ones.

**Tip:** Click the play button next to a voice to hear a sample.

### Step 4: Adjust Settings

* **Speed:** How fast the voice speaks (0.5x to 2x)
* **Model:** Different AI models for varying quality

### Step 5: Generate

1. Click **Generate**
2. Wait a few seconds
3. Play your audio to review

Happy with it? Download and use it. Need changes? Adjust and regenerate.

## Choosing the Right Voice

### Voice Characteristics

| Characteristic | Options                                 |
| -------------- | --------------------------------------- |
| **Gender**     | Male, Female, Neutral                   |
| **Accent**     | American, British, Australian, etc.     |
| **Age**        | Young adult, Middle-aged, Mature        |
| **Tone**       | Professional, Friendly, Energetic, Calm |

### Matching Voice to Content

| Content Type        | Recommended Voice                  |
| ------------------- | ---------------------------------- |
| Corporate training  | Professional, clear, moderate pace |
| Marketing video     | Energetic, engaging, friendly      |
| Meditation/wellness | Calm, slow, soothing               |
| Explainer video     | Clear, friendly, moderate pace     |
| Phone system        | Professional, neutral, clear       |

## Writing for Audio

Text written for reading doesn't always sound natural when spoken. Here are tips for better audio:

### Write Conversationally

**Avoid — Written style:**
"The implementation of our solution facilitates operational efficiency."

**Better — Audio style:**
"Our solution makes your work faster and easier."

### Use Punctuation for Pacing

Punctuation affects how the AI reads your text:

* **Commas:** Short pause
* **Periods:** Medium pause
* **Dashes (—):** Dramatic pause
* **Ellipses (...):** Thoughtful pause

**Example:**

```
This is important... really important.
And here's why — our customers love it.
```

### Spell Out Abbreviations

The AI might mispronounce abbreviations:

* **Avoid:** "CEO" might be read as "see-ee-oh"
* **Better:** "Chief Executive Officer" or write phonetically

### Add Phonetic Hints for Unusual Words

For names or technical terms:

```
John Scieszka (pronounced "SHESS-ka")
```

## Adjusting Speed

The speed setting affects how fast the voice speaks:

| Speed     | Best For                                   |
| --------- | ------------------------------------------ |
| **0.75x** | Detailed explanations, emphasis, tutorials |
| **1.0x**  | Normal conversation, most content          |
| **1.25x** | Quick announcements, summaries             |
| **1.5x**  | Fast-paced content, short messages         |

**Tip:** Most professional voiceovers use 0.9x to 1.1x speed.

## Managing Your Audio Files

### Viewing Your Library

All generated audio is saved automatically:

1. Go to **Audio**
2. Browse your list
3. Click play to preview any file

### Playing Audio

1. Click on an audio file
2. Use the play controls
3. Adjust volume as needed

### Downloading Audio

1. Click on an audio file
2. Click **Download**
3. File saves as MP3 (most common) or other format

### Deleting Audio

1. Click the **three-dot menu** (⋮) on the audio file
2. Select **Delete**
3. Confirm

## Tips for Better Audio

### 1. Break Long Text into Sections

Instead of one long script, create separate audio files for:

* Intro
* Main content sections
* Call to action
* Outro

This makes editing easier and allows mixing voices.

### 2. Test Before Full Production

Generate a short sample first to verify:

* The voice sounds right
* Speed is appropriate
* Pronunciation is correct
* Tone matches your content

### 3. Review and Re-Generate

Listen critically:

* Does it sound natural?
* Are pauses in the right places?
* Is the emphasis correct?

Adjust text or settings and regenerate as needed.

### 4. Consider Your Audience

* **International audience:** Choose a neutral accent
* **Technical content:** Slower speed for comprehension
* **Young audience:** More energetic voices

## Audio Length Limits

Most models have character limits per generation:

| Model    | Character Limit    |
| -------- | ------------------ |
| Standard | 5,000 characters   |
| Advanced | 10,000+ characters |

For longer content, split into multiple audio files.

## Using Audio in Projects

### Adding to Videos

1. Download your audio file
2. Import into your video editor
3. Sync with visuals

### Podcasts

1. Generate intro/outro clips
2. Create spoken content sections
3. Combine in your audio editor

### Phone Systems (IVR)

1. Generate short, clear prompts
2. Export in required format
3. Upload to your phone system

## Custom Voices

Want a unique voice that sounds like you or represents your brand? See [Voices](/engage/content-engine/voices) to learn about voice cloning.

## Frequently Asked Questions

**Q: How long does audio generation take?**
A: Usually 5-15 seconds for short clips, up to a minute for longer content.

**Q: What format are audio files?**
A: Typically MP3. Some systems offer additional formats.

**Q: Can I use these for commercial purposes?**
A: Yes, check your subscription terms for specific commercial use rights.

**Q: Why does the voice mispronounce a word?**
A: Try spelling it phonetically or differently. AI learns from common pronunciations.

**Q: Can I make the voice sound emotional?**
A: To some extent. Use punctuation and word choice to guide emotion. For more control, consider using custom [Voices](/engage/content-engine/voices).

## Next Steps

* [Create Custom Voices](/engage/content-engine/voices)
* [Generate Videos](/engage/content-engine/video)
* [Back to Content Engine](/engage/content-engine)


# Brand Images
Source: https://docs.parallellabs.app/engage/content-engine/brand-images

Save reusable brand styles and frame templates so generated images stay on-brand.

Brand Images let you create reusable image collections with consistent styling for your AI-generated content. Define your brand's visual identity once, then use it across all your campaigns.

## What Are Brand Images?

Brand Images are collections that include:

* **Reference Images:** Logos, product photos, icons, and other brand assets
* **Style Prompts:** AI-generated descriptions that maintain visual consistency
* **AI Models:** Preferred image generation models for your brand
* **Frame Templates:** Transparent overlays to brand your generated images

## Creating a Brand Image Collection

### Step 1: Open Brand Images

1. Select your company from the sidebar
2. Click **Brand Images** (under Content Tools)
3. Click **New Brand Image**

### Step 2: Basic Information

* **Name:** A descriptive name (e.g., "Company Marketing", "Product Line A")
* **Description:** Notes about when to use this collection

### Step 3: Add Reference Images

Upload images that represent your brand:

1. Click **Upload Image** in the Images section
2. Select an image from your computer
3. For each image, set:
   * **Type:** Logo, Product, Hero, Icon, Banner, Social Media, Character, or Other
   * **Description:** How this image should be used

**Tip:** Add your logo and key product images so the AI can reference them.

### Step 4: Configure AI Models

Select which AI image generation models to use with this brand:

1. Check the models you want to enable
2. Different models produce different styles
3. You can select multiple models

### Step 5: Generate a Style Prompt

Create a consistent style description for all generated images:

1. Enter a description of your desired style in the text field
2. Click **Generate Style Prompt** (magic wand icon)
3. The AI creates a detailed Midjourney-style prompt
4. Edit as needed

**Example style prompt:**

```
professional aesthetic of a modern tech platform, 
soft pastel color palette with warm tones, 
gentle and playful textures, diffused natural lighting, 
balanced composition with centered focus, 
warm inviting mood --ar 16:9 --style raw --v 6
```

## Creating Frame Templates with AI

Frame templates are transparent PNG overlays that brand your AI-generated images. Use the AI Frame Builder to create them.

### Opening the Frame Builder

1. In the Brand Image dialog, scroll to **Templates**
2. Click **Create with AI**

### Building Your Frame

1. **Select Aspect Ratio:** Choose from presets like Square (1:1), Instagram Story (9:16), Landscape (16:9), etc.
2. **Name Your Frame:** Enter a descriptive name
3. **Describe Your Design:** Chat with the AI to build your frame

**Example prompts:**

* "Create a frame with our company name in the bottom-right corner with a subtle gradient"
* "Add a 'NEW' badge in the top-left corner"
* "Make a professional watermark with slight transparency"

### Frame Design Tips

* **Keep the center empty:** The main image shows through transparent areas
* **Use edges and corners:** Place logos, text, and decorations around the borders
* **Consider contrast:** Use shadows and outlines so elements are readable over any image
* **Test different backgrounds:** The preview shows a checkered pattern for transparency

### Exporting Your Frame

1. Click **Export & Add Template**
2. The frame is saved as a transparent PNG
3. It's automatically added to your templates list

## Using Brand Images in Campaigns

When generating images for content:

1. Select your Brand Image collection
2. The AI uses your:
   * Style prompt for consistent visuals
   * Reference images for context
   * Preferred AI models
3. Apply frame templates to finished images

## Managing Brand Images

### Editing a Collection

1. Click the **three-dot menu** (⋮) on any row
2. Select **Edit**
3. Make your changes
4. Click **Save**

### Deleting a Collection

1. Click the **three-dot menu** (⋮)
2. Select **Delete**
3. Confirm deletion

**Note:** Deleting a brand image collection doesn't delete the actual images, only the collection configuration.

## Best Practices

### 1. Create Purpose-Specific Collections

Instead of one generic collection, create specific ones:

* "Social Media Marketing"
* "Product Announcements"
* "Blog Post Headers"

### 2. Keep Style Prompts Updated

As your brand evolves, update your style prompts to match.

### 3. Use Descriptive Image Labels

When adding reference images, write clear descriptions so the AI knows how to use them:

* **Better:** "Primary logo for dark backgrounds, use in bottom-right corner"
* **Avoid:** "logo2.png"

### 4. Test Frame Templates

Before using frames in production:

1. Generate a test image
2. Apply your frame template
3. Verify it looks good on various image types

## Troubleshooting

### Frame has padding when overlaid

If your frame appears smaller than the image:

* Ensure the frame dimensions match your target aspect ratio
* Regenerate the frame with the AI Frame Builder

### Style prompt not working

If generated images don't match your style:

* Make the style prompt more specific
* Include concrete details about colors, lighting, and mood
* Regenerate the style prompt with more context

### Reference images not appearing

* Check that images uploaded successfully
* Ensure images have descriptions
* Try re-uploading the image


# Brand Voice
Source: https://docs.parallellabs.app/engage/content-engine/brand-voice

Define how your content should sound, so everything you generate reads like it came from the same company.

Your brand voice is the personality, tone, and style of your writing. Setting one means you stop editing the same things out of every AI draft — the corporate buzzwords, the wrong level of formality, the phrases you'd never actually use.

## Creating a Brand Voice

1. Select your company from the sidebar
2. Go to **Content Engine** → **Brand Voice**
3. Click **New Brand Voice** — or **Generate with AI** to infer a voice from your existing material
4. Define:
   * **Personality** — professional, casual, playful, authoritative?
   * **Tone** — friendly, formal, conversational?
   * **Writing style** — simple, technical, storytelling?
   * **Words to use** — terminology and phrases you like
   * **Words to avoid** — language that doesn't fit your brand
5. Save

The "words to avoid" list does more work than people expect. It's the difference between drafts you edit and drafts you publish.

## Example

```
Personality: Friendly expert
Tone: Warm but professional
Style: Clear and direct, avoid jargon

DO use:
- "You" and "your" to speak directly to readers
- Simple explanations for complex topics
- Real examples and stories

DON'T use:
- Corporate buzzwords like "synergy" or "leverage"
- Overly casual language like "gonna" or "wanna"
- Fear-based marketing tactics
```

## Managing Brand Voices

The list shows each voice's **Name**, **Industry**, **Target Audience**, and when it was created and updated. Use a voice's menu to **View Details**, **Edit**, or **Delete**.

You can keep more than one — a formal voice for whitepapers and a looser one for social, for instance — and point different [campaigns](/engage/content-engine/content-campaigns) at different voices.

## Where It's Used

A brand voice applies anywhere the AI writes for you: generated [content](/engage/content-engine/content), [campaign](/engage/content-engine/content-campaigns) output, and drafts produced from an [opportunity](/engage/content-engine/seo-opportunities) brief.

## Tips

1. **Show, don't just label.** "Warm but professional" means less than two example sentences demonstrating it.
2. **Be specific about what to avoid.** Naming the exact words you hate works better than "avoid jargon."
3. **Derive it from real writing.** Point **Generate with AI** at material you already like, then refine.
4. **Keep it short enough to be followed.** A voice guide nobody can hold in their head is a voice guide the output won't reflect either.

## Next Steps

* [Define your Ideal Customer Profile](/engage/content-engine/icp)
* [Plan a Content Strategy](/engage/content-engine/content-strategy)
* [Set up Brand Images](/engage/content-engine/brand-images)


# Canvas
Source: https://docs.parallellabs.app/engage/content-engine/canvas

Design social graphics, story art, quote cards, and link previews with an AI design assistant.

[Images](/engage/content-engine/images) generates a picture from a prompt. **Canvas** designs a *layout* — text, shapes, and imagery composed at an exact size for a specific placement, which you export as a PNG (or an MP4, if it's animated).

Use it for the graphics that carry words: quote cards, announcement posts, story art, link previews.

## Creating a Canvas

1. Select your company from the sidebar
2. Go to **Content Engine** → **Canvas**
3. Click **New Canvas**
4. Give it a **Canvas Name** — for example, "Product Launch Story"
5. Pick a **Size**
6. Turn on **Animated** if you want motion
7. Click create

## Sizes

Pick the size for where it will be posted — designing at the target size avoids the cropping that ruins text.

| Size                       | Dimensions  | Where it's used              |
| -------------------------- | ----------- | ---------------------------- |
| **Square (1:1)**           | 1080 × 1080 | Instagram and LinkedIn feed  |
| **Portrait (4:5)**         | 1080 × 1350 | Instagram feed, taller crop  |
| **Story (9:16)**           | 1080 × 1920 | Stories, Reels, TikTok       |
| **Landscape (1.91:1)**     | 1200 × 630  | Link previews and Open Graph |
| **Landscape (16:9)**       | 1280 × 720  | YouTube thumbnails           |
| **X/Twitter Post (16:9)**  | 1200 × 675  | X posts                      |
| **LinkedIn Post (1.91:1)** | 1200 × 627  | LinkedIn posts               |
| **Facebook Post (1.91:1)** | 1200 × 628  | Facebook posts               |

## Designing with the Assistant

The builder is a chat on the left and a live preview on the right. Describe what you want and the assistant designs it; keep talking to refine it.

Work in small, specific steps. "Make the headline bigger and move the logo to the bottom right" gets you further than restating the whole brief each time. **Refresh Preview** re-renders if the preview drifts out of date.

You can change the **Size** from the top bar, and rename the canvas at any time.

## Exporting

Click **Export PNG** to render the finished canvas. Animated canvases export as a looping video with **Export MP4** instead.

The export is a normal file — attach it to [content](/engage/content-engine/content), post it, or use it anywhere else.

## Tips

1. **Choose the size before designing.** Resizing afterwards reflows the layout.
2. **Keep the text short.** Social graphics are read at a glance and at thumbnail scale.
3. **Iterate one change at a time**, so you can tell what actually improved it.
4. **Use animation sparingly** — for stories and reels, where motion is expected.

## Next Steps

* [Generate images from prompts](/engage/content-engine/images)
* [Set up Brand Images for consistency](/engage/content-engine/brand-images)
* [Attach visuals to your content](/engage/content-engine/content)


# Content
Source: https://docs.parallellabs.app/engage/content-engine/content

Every piece you've created, in a list or a calendar, with approvals, scheduling, and publishing.

**Content** is the working surface of the Content Engine. Everything generated by a [campaign](/engage/content-engine/content-campaigns), drafted from an [opportunity](/engage/content-engine/seo-opportunities), or created by hand shows up here to be reviewed, approved, scheduled, and published.

## Creating Content

Click **Create from Idea**, describe what you want, and the AI drafts it using your [ICP](/engage/content-engine/icp), [strategy](/engage/content-engine/content-strategy), and [brand voice](/engage/content-engine/brand-voice) if you've set them up.

Content also arrives here automatically from [campaigns](/engage/content-engine/content-campaigns) that are running on a schedule.

### Writing a good prompt

**Avoid:** "Write about marketing"

**Better:** "Write a 500-word blog post about 5 email marketing mistakes small businesses make, with tips to fix each one"

The more specific the ask — length, format, audience, angle — the less editing the draft needs.

## List and Calendar Views

Toggle between **List** and **Calendar**.

* **List** is for working through content — filtering, bulk actions, approvals.
* **Calendar** is for seeing publishing cadence — where the gaps and the pile-ups are.

## Finding Things

The filters along the top narrow the list:

| Filter               | Options                                            |
| -------------------- | -------------------------------------------------- |
| **Search**           | Free text                                          |
| **Campaign**         | Which campaign produced it                         |
| **Type**             | Blog post, social post, marketing email, and so on |
| **Status**           | Ideated, Draft, Published, Scheduled               |
| **Approval**         | Unassigned, pending, in review, approved, rejected |
| **Funnel Stage**     | TOFU, MOFU, BOFU                                   |
| **Date Filter Type** | Filter by creation date or publish date            |
| **Sort By**          | Date, **Top Views**, or **Fewest Views**           |

**Clear filters** resets everything at once.

Sorting by **Top Views** is the quickest way to see what's actually landing; **Fewest Views** shows you what to stop making.

## Status

| Status         | Meaning                           |
| -------------- | --------------------------------- |
| **Ideated**    | An idea captured, not yet written |
| **Generating** | Being written right now           |
| **Draft**      | Written, not published            |
| **Scheduled**  | Queued to publish at a set time   |
| **Published**  | Live                              |

## Publishing

Open a piece's menu and you'll see the actions that apply to it:

* **Publish or Schedule…** — for blog, LinkedIn, Facebook, and Instagram posts in draft or scheduled state
* **Publish as Marketing Campaign…** — turns a marketing email into a [broadcast send](/engage/marketing-campaigns)
* **Set as Published** — mark a draft live if you published it elsewhere yourself
* **Move to Draft** — pull a scheduled piece back out of the queue

Publishing to a social or blog destination requires the matching [integration](/settings/integrations) to be connected.

## Approvals

Content can be routed through a person before it goes out:

1. Use **Assign to Collaborator** on a piece
2. The assignee sees it and can **Review Content**
3. Its approval state moves through pending, in review, and then approved or rejected

Filter by **Approval** to find everything waiting on you — or everything you're still waiting on someone else for. Assignees must be [collaborators](/settings/collaborators) on the company.

## Bulk Actions

Tick several rows and you can **Schedule** or **Delete** them together, or **Clear selection**. **Export/Import** moves content in and out in bulk.

## Tips

1. **Always edit AI drafts.** They're a strong first pass, not a finished piece — add the specifics only you know.
2. **Review before scheduling.** Scheduling unread content means publishing unread content.
3. **Use funnel filters to spot gaps.** All-BOFU content means nothing is bringing new readers in.
4. **Let the views columns guide you.** Repeat what works instead of guessing.

## Next Steps

* [Set up a recurring campaign](/engage/content-engine/content-campaigns)
* [Find topics with real demand](/engage/content-engine/seo-opportunities)
* [Generate images to go with it](/engage/content-engine/images)


# Campaigns
Source: https://docs.parallellabs.app/engage/content-engine/content-campaigns

Recurring content programs that generate posts for you — on demand, or on a schedule.

A content campaign is a standing instruction to produce content. You describe the program once — what it's about, who it's for, how it should sound, how often it runs — and it generates pieces into your [Content](/engage/content-engine/content) list.

> **Not to be confused with [Marketing Campaigns](/engage/marketing-campaigns)**, which send a broadcast email. A *content* campaign produces content; a *marketing* campaign sends mail.

## Creating a Campaign

1. Select your company from the sidebar
2. Go to **Content Engine** → **Campaigns**
3. Click **New Campaign**
4. Fill in the campaign:
   * **Name**
   * **Campaign Plan** — what this program is about and what it should produce
   * **Tone** — Professional, Casual, Friendly, or Authoritative
   * **Enable Web Research** — let it research current information while writing
   * **Enable Knowledge Base** — let it draw on your [knowledge base](/overview/knowledge-base)
   * **Auto-Run & Publishing Frequency** — see below
   * Optionally link an **[Ideal Customer Profile](/engage/content-engine/icp)**, **[Content Strategy](/engage/content-engine/content-strategy)**, **[Brand Voice](/engage/content-engine/brand-voice)**, and **[Brand Image](/engage/content-engine/brand-images)**
5. Save

Linking the brand records is what makes campaign output specific to your business rather than generic.

## Frequency

| Setting     | What happens                                                          |
| ----------- | --------------------------------------------------------------------- |
| **Manual**  | Nothing runs on its own — you click **Trigger** when you want content |
| **Daily**   | Runs every day                                                        |
| **Weekly**  | Runs every 7 days                                                     |
| **Monthly** | Runs every 30 days                                                    |

Campaigns default to **Manual**. A campaign left on Manual will never produce anything by itself — if you expected content to appear and it hasn't, check this setting first.

## Managing Campaigns

The list shows each campaign's **Name**, **Frequency**, **Status**, **Content Count**, and when it was created and updated.

Status tells you how the program is doing:

| Status      | Meaning              |
| ----------- | -------------------- |
| **Healthy** | Running normally     |
| **Running** | Generating right now |
| **Error**   | The last run failed  |

From a campaign's menu you can **Trigger** it to run now, **Edit** it, or **Delete** it.

## Reviewing What It Produces

Generated pieces land in [Content](/engage/content-engine/content) as drafts, not published posts. Nothing goes live without you — filter the content list by campaign to review a program's output, then publish or schedule the pieces you want.

## Tips

1. **Write the plan like a brief.** "Weekly post on warehouse automation for ops managers, one practical takeaway each" produces far better output than "content about our industry."
2. **Start on Manual.** Trigger it a couple of times and read the results before putting it on a schedule.
3. **Enable web research for anything time-sensitive**, and leave it off for evergreen material.
4. **Watch the Error status.** A campaign that has been failing quietly has produced nothing for however long it's been broken.

## Next Steps

* [Review and publish your content](/engage/content-engine/content)
* [Set up a Brand Voice](/engage/content-engine/brand-voice)
* [Find topics worth covering](/engage/content-engine/seo-opportunities)


# Content Strategy
Source: https://docs.parallellabs.app/engage/content-engine/content-strategy

Decide what topics you'll own, and map them across the stages of your funnel.

A content strategy defines **what you'll create and why**. It's built from **content pillars** — the recurring themes you want to be known for — each mapped to a stage of the buyer's journey.

## Creating a Strategy

1. Select your company from the sidebar
2. Go to **Content Engine** → **Content Strategy**
3. Click **New Strategy** — or **Generate with AI** to draft pillars from your business and [ICP](/engage/content-engine/icp)
4. Define content pillars for each stage of your funnel
5. Save

## The Funnel Stages

Content is mapped to three stages, and the AI adjusts its messaging depending on which one a piece targets.

| Stage                                      | Reader's mindset            | Content that fits                                             |
| ------------------------------------------ | --------------------------- | ------------------------------------------------------------- |
| **TOFU** — Top of Funnel, Awareness        | "I have a problem"          | Educational content, industry trends, general problem-solving |
| **MOFU** — Middle of Funnel, Consideration | "What are my options?"      | How-to guides, comparisons, case studies                      |
| **BOFU** — Bottom of Funnel, Decision      | "Is this the right choice?" | Product demos, testimonials, pricing and ROI content          |

A healthy strategy covers all three. Most teams over-produce BOFU content — which only speaks to people already close to buying — and under-produce the TOFU content that brings new readers in.

## Managing Strategies

The list shows each strategy's **Name**, its **Content Pillars**, its **Funnel Coverage**, and when it was created and updated. Funnel coverage is the quick check: it tells you whether you've actually planned for all three stages or just one.

Click a strategy to open it, or use its menu to **Edit** or **Delete**.

## Using a Strategy

When you create a piece of [content](/engage/content-engine/content), you can set which funnel stage it's for, and the AI adjusts the messaging accordingly. You can also filter your content list by funnel stage to see where the gaps are.

[Campaigns](/engage/content-engine/content-campaigns) can be linked to a strategy so everything a campaign generates stays on-theme.

## Tips

1. **Three to five pillars is plenty.** More than that and you're not focused, you're just listing topics.
2. **Pillars should be durable.** A pillar you can write about for a year is a pillar; a pillar you exhaust in three posts is a topic.
3. **Check coverage, not volume.** Twelve BOFU posts and no TOFU means nobody new is finding you.
4. **Let [Opportunities](/engage/content-engine/seo-opportunities) inform it.** Real search demand is better evidence than intuition about what to cover.

## Next Steps

* [Define your Ideal Customer Profile](/engage/content-engine/icp)
* [Establish a Brand Voice](/engage/content-engine/brand-voice)
* [Find topics worth writing about](/engage/content-engine/seo-opportunities)


# Ideal Customer Profile
Source: https://docs.parallellabs.app/engage/content-engine/icp

Describe who you're creating content for, so everything generated speaks to the right person.

An **Ideal Customer Profile** (ICP) is a description of the person you're trying to reach. It's the first thing the AI reads when it writes for you — without one, it has to guess at your audience and writes generically.

## Creating a Profile

1. Select your company from the sidebar
2. Go to **Content Engine** → **Ideal Customer Profile**
3. Click **New Profile** — or **Generate with AI** to draft one from what the platform already knows about your business
4. Fill in details about your target customer:
   * Demographics — age, location, job title
   * Pain points — what problems do they have?
   * Goals — what are they trying to achieve?
   * How they make decisions
   * Where they consume content
   * Industry, budget range, and location
5. Save

**Generate with AI** is usually the fastest start: let it draft a profile, then correct the parts it got wrong. Editing a draft is quicker than filling a blank form.

## Managing Profiles

The list shows each profile's **Name**, **Description**, **Industry**, **Budget Range**, **Location**, and when it was created and updated.

From a profile's menu you can **View Details**, **Edit**, or **Delete** it.

You can keep several profiles — one per segment you sell to — and point different [campaigns](/engage/content-engine/content-campaigns) at different profiles.

## Why It Matters

When the AI generates content, it uses your profile to:

* Choose the right tone and vocabulary
* Address pain points that are actually relevant
* Emphasize the benefits that matter to that buyer
* Pitch at the right level of technical detail

The difference is concrete: content written for "operations managers at 50–200 person logistics firms" reads very differently from content written for nobody in particular.

## Where Profiles Are Used

* **[Content Strategy](/engage/content-engine/content-strategy)** — pillars are planned around a profile
* **[Campaigns](/engage/content-engine/content-campaigns)** — a campaign can be linked to a profile
* **[Opportunities](/engage/content-engine/seo-opportunities)** — research projects target an ICP
* **[Smart Lists](/find-enrich/smart-lists)** — AI scoring can rate rows against your ICP

## Tips

1. **Be specific.** "Small business owners" is too broad to change what the AI writes. "Owner-operators of 5–20 vehicle trucking firms" is usable.
2. **Write pain points in their words**, not your marketing language.
3. **One profile per segment.** If you sell to two very different buyers, make two profiles rather than averaging them into one.
4. **Revisit it.** When your positioning changes, the profile should change too.

## Next Steps

* [Plan a Content Strategy](/engage/content-engine/content-strategy)
* [Establish a Brand Voice](/engage/content-engine/brand-voice)
* [Create content](/engage/content-engine/content)


# Images
Source: https://docs.parallellabs.app/engage/content-engine/images

Generate and edit images from text prompts.

Create custom images using AI text-to-image technology. Perfect for social media graphics, blog illustrations, marketing materials, and more.

## What Can You Create?

* Marketing graphics and banners
* Social media images
* Blog post illustrations
* Product mockups
* Abstract backgrounds
* Conceptual artwork
* And more!

## Generating Your First Image

### Step 1: Open Image Generation

1. Select your company from the sidebar
2. Click **Images** (under the Content Engine menu)
3. Click **Generate Image** (or the **+** button)

### Step 2: Write Your Prompt

Describe the image you want. The more detail, the better:

**Basic prompt:**

```
A modern office workspace
```

**Better prompt:**

```
A bright, modern office workspace with large windows,
a wooden desk with a laptop, green plants, and soft
natural lighting. Minimalist style, warm colors.
```

### Step 3: Choose Settings

* **Model:** Different AI models create different styles
* **Size:** Square (1024x1024), landscape, or portrait
* **Style:** Vivid (bold colors) or Natural (realistic)
* **Quality:** Standard or HD

### Step 4: Generate

1. Click **Generate**
2. Wait a few seconds
3. Your image appears!

Not quite right? Adjust your prompt and generate again.

## Writing Great Prompts

### Include These Elements

| Element         | Example                   |
| --------------- | ------------------------- |
| **Subject**     | "A golden retriever"      |
| **Action/Pose** | "running through a field" |
| **Setting**     | "in a sunny meadow"       |
| **Style**       | "photorealistic style"    |
| **Lighting**    | "golden hour lighting"    |
| **Mood**        | "joyful and energetic"    |
| **Colors**      | "warm autumn colors"      |

### Example Prompts

**For a business blog:**

```
Professional business team having a productive meeting
in a modern glass conference room. Diverse group of
people, natural lighting, corporate but friendly
atmosphere. Clean, professional photography style.
```

**For social media:**

```
Flat lay photography of a coffee cup, notebook,
and succulent plant on a white marble desk.
Minimalist aesthetic, soft shadows, Instagram-worthy
product photography style.
```

**For abstract backgrounds:**

```
Abstract flowing gradient background in soft purple
and blue tones. Smooth, professional, suitable for
website header. Modern and clean design.
```

## Understanding Image Settings

### Models

Different models excel at different styles:

| Model                           | Best For                                         |
| ------------------------------- | ------------------------------------------------ |
| **GPT Image 1 / 1.5 / 2**       | General purpose, high quality, and precise edits |
| **FLUX.1 Kontext**              | Controllable generation and editing              |
| **Leonardo (Phoenix, Kino XL)** | Strong prompt adherence and text rendering       |
| **Flux Schnell**                | Fast drafts                                      |
| **Nano Banana (Google)**        | Google's Gemini image model                      |

The exact model list depends on your plan. Try a few to see which works best for your needs. Some models (like the GPT Image family and Leonardo) also accept **reference images** so you can guide the result or edit an existing image.

### Sizes

| Size          | Dimensions | Best For                 |
| ------------- | ---------- | ------------------------ |
| **Square**    | 1024x1024  | Social media, thumbnails |
| **Landscape** | 1792x1024  | Blog headers, banners    |
| **Portrait**  | 1024x1792  | Stories, Pinterest       |

### Style Options

* **Vivid:** Bold, dramatic colors and high contrast
* **Natural:** More realistic, subdued tones

### Quality

* **Standard:** Faster, good for drafts and testing
* **HD:** Higher detail, best for final outputs

## Using Reference Images

Some models let you use existing images as inspiration:

### How to Use References

1. Click **Add Reference Image**
2. Upload an image or select from your library
3. Describe how you want to modify or combine elements
4. Generate

**Example:**
Upload your product photo and prompt: "Create a lifestyle scene with this product on a modern kitchen counter with morning coffee and fresh fruit."

## Managing Your Images

### Viewing Your Library

All generated images are saved automatically:

1. Go to **Images**
2. Browse your gallery
3. Click on any image to view full size

### Downloading Images

1. Click on an image
2. Click **Download**
3. Image saves to your computer

### Copying to Clipboard

1. Click on an image
2. Click **Copy**
3. Paste anywhere (documents, social media, etc.)

### Deleting Images

1. Click the **three-dot menu** (⋮) on an image
2. Select **Delete**
3. Confirm

## Tips for Better Images

### 1. Be Specific

Generic prompts get generic results. Add details about:

* Colors and color scheme
* Lighting and atmosphere
* Composition and framing
* Style and artistic approach

### 2. Iterate

Your first image is rarely perfect. Try:

* Adjusting your prompt slightly
* Changing the style setting
* Using a different model
* Adding or removing details

### 3. Use Style Keywords

Add these to your prompts:

* "Professional photography"
* "Minimalist design"
* "Flat illustration"
* "Watercolor painting"
* "3D render"
* "Vintage aesthetic"

### 4. Avoid Common Issues

**Hands and text:** AI often struggles with these. If needed, edit them out or plan to avoid them.

**Faces:** For professional use, verify faces look natural. Consider abstract or distant subjects.

**Brands and logos:** AI shouldn't reproduce copyrighted content. Create original designs instead.

## Image Credits and Usage

### Credit Consumption

Image generation uses credits from your account. Typically:

* Standard images: Moderate credits
* HD images: More credits
* Reference-based images: More credits

Check your [Profile](/account/profile) to monitor usage.

### Usage Rights

Images you generate are yours to use for:

* Marketing and advertising
* Social media
* Website content
* Internal documents

Always check your subscription terms for commercial use details.

## Frequently Asked Questions

**Q: How long does image generation take?**
A: Usually 10-30 seconds depending on settings and demand.

**Q: Why doesn't my image match what I described?**
A: Try adding more specific details, rephrasing, or breaking complex ideas into simpler prompts.

**Q: Can I edit generated images?**
A: Currently, the platform generates images but doesn't include editing. Download and use your preferred image editor.

**Q: Are there content restrictions?**
A: Yes. The AI won't generate inappropriate, harmful, or copyrighted content.

**Q: Can I generate images of real people?**
A: No. Creating images of specific real people without consent isn't supported for privacy and ethical reasons.

## Next Steps

* [Create Audio Content](/engage/content-engine/audio)
* [Generate Videos](/engage/content-engine/video)
* [Back to Content Engine](/engage/content-engine)


# Domain Overview
Source: https://docs.parallellabs.app/engage/content-engine/seo-domain-overview

Authority, traffic, and backlinks for your domain and your competitors'.

Domain Overview tells you how strong a website is in search. Use it on your own domain to know where you stand, and on competitors' to know who you're realistically up against.

## Looking Up a Domain

1. Select your company from the sidebar
2. Go to **Content Engine** → **Domain Overview**
3. Enter a **Domain** — or click **Use My Domain**
4. Optionally add **Competitor domains**, comma separated
5. Review the comparison

Your own domain is marked with a **You** chip. Cached results carry a **From cache** chip.

## What the Metrics Mean

| Metric                | What it tells you                                           |
| --------------------- | ----------------------------------------------------------- |
| **Domain Rating**     | Overall authority — how much weight this site's pages carry |
| **Monthly Traffic**   | Estimated visits from search                                |
| **Referring Domains** | How many distinct sites link to it                          |
| **Backlinks**         | Total inbound links                                         |

**Referring Domains** is usually the more honest signal of the two link metrics — a thousand backlinks from three sites means far less than a hundred from a hundred.

## Using It Competitively

Compare your domain against three or four real competitors. The gap in Domain Rating tells you how ambitious your keyword targets can be: if a competitor is far stronger, going after their head terms directly will be slow, and you're better served by more specific topics where authority matters less.

## Credits

Each **uncached domain** costs **25 credits**. Cached lookups are free, and **Refresh Data** re-purchases at full cost.

Adding four competitor domains at once means four lookups, so add the ones you'll actually act on.

## Tips

1. **Compare against real competitors**, not aspirational ones.
2. **Track it over time.** A single reading is a snapshot; the direction of travel is the useful part.
3. **Don't chase Domain Rating directly.** It's a symptom of good content and links, not a lever.

## Next Steps

* [Research keywords](/engage/content-engine/seo-keyword-research)
* [Audit your own site](/engage/content-engine/seo-site-audit)
* [Find content opportunities](/engage/content-engine/seo-opportunities)


# Keyword Research
Source: https://docs.parallellabs.app/engage/content-engine/seo-keyword-research

Search demand, cost per click, and competition for any keyword.

Keyword Research answers the basic question behind every content decision: **is anyone actually searching for this?** Enter a keyword and get back its search volume along with related terms you may not have thought of.

## Running a Search

1. Select your company from the sidebar
2. Go to **Content Engine** → **Keyword Research**
3. Enter a **Keyword** — for example, "project management software"
4. Choose how to **Sort by**: Volume, CPC, Competition, or Keyword
5. Click **Search**

You get the keyword you typed — marked with a **Seed** chip — plus related keywords, each with its own metrics.

## Reading the Results

| Column          | What it tells you                                              |
| --------------- | -------------------------------------------------------------- |
| **Volume**      | How many searches per month                                    |
| **CPC**         | What advertisers pay per click — a proxy for commercial intent |
| **Competition** | How hard the term is to compete for                            |
| **Trend**       | Whether interest is rising or falling                          |
| **Saved**       | Whether you've starred it                                      |

High CPC with moderate volume often beats high volume with no CPC: it means the traffic is worth money to someone.

## Saving Keywords

Tick the keywords worth keeping and click **Save Keywords**. Saved keywords go to [Keywords](/engage/content-engine/seo-keywords), where they're stored with their metrics for later.

## Credits

A lookup costs **25 credits**, and results are cached — repeating a search you've already run is free, and the results line tells you which happened (`25 credits used` or `served from cache, no credits used`).

**Refresh Data** deliberately re-purchases metrics at full cost. Use it when you think the numbers have gone stale, not as a matter of routine.

## Tips

1. **Search the phrase your customer would type**, not your product's internal name.
2. **Mine the related keywords.** The seed term is rarely the best one — the list around it is the point.
3. **Read volume and competition together.** A low-volume term you can actually rank for beats a high-volume term you can't.
4. **Let the cache work.** Re-running a search costs nothing; refreshing costs 25 credits.

## Next Steps

* [Review your saved keywords](/engage/content-engine/seo-keywords)
* [Check domain authority](/engage/content-engine/seo-domain-overview)
* [Find ranked content opportunities](/engage/content-engine/seo-opportunities)


# Keywords
Source: https://docs.parallellabs.app/engage/content-engine/seo-keywords

Every keyword from your research runs and lookups, with the ones you've starred kept together.

Where [Keyword Research](/engage/content-engine/seo-keyword-research) buys new data, **Keywords** is the store of what you've already looked up. Nothing here costs credits — it's your own history, kept with its metrics.

## Browsing Keywords

1. Select your company from the sidebar
2. Go to **Content Engine** → **Keywords**

Switch between **All** and **Saved** to see everything you've ever looked up, or just the keywords you've starred.

Narrow the list with:

* **Search keywords** — free text
* **Sort by** — reorder by the metric you care about
* **Tag** — filter to a single tag

## The Table

| Column          | What it shows                |
| --------------- | ---------------------------- |
| **Keyword**     | The term                     |
| **Volume**      | Monthly searches             |
| **CPC**         | Cost per click               |
| **Competition** | How contested it is          |
| **Tags**        | Your own labels              |
| **Saved**       | Star to keep it in **Saved** |

## Using Tags

Tags are how this stays usable once you have hundreds of keywords. Group them the way you'll actually work — by content pillar, by funnel stage, by campaign, or by quarter — then filter to one tag when you're planning what to write.

## Starring

Starring moves a keyword into **Saved**. Use it as a shortlist: terms you've decided are worth targeting, separated from everything you merely looked at once.

## Tips

1. **Star sparingly.** A shortlist of 30 is useful; a shortlist of 300 is a second list of everything.
2. **Tag as you save**, not later — retroactive tagging rarely happens.
3. **Revisit before each planning cycle.** Saved keywords are a ready-made backlog of topics.

## Next Steps

* [Research more keywords](/engage/content-engine/seo-keyword-research)
* [Turn keywords into content](/engage/content-engine/content)
* [See ranked opportunities](/engage/content-engine/seo-opportunities)


# Opportunities
Source: https://docs.parallellabs.app/engage/content-engine/seo-opportunities

Ranked content opportunities, each with real search volume and a ready-to-write brief.

The other SEO pages give you raw numbers. **Opportunities** does the analysis: it studies your search landscape, your competitors, and what AI assistants say when asked about your space, then hands you a ranked list of things worth writing — each with a brief you can generate from directly.

## Why AEO Matters

Search isn't only Google any more. When someone asks ChatGPT or Google's AI which tools to use, something gets cited — and it may not be you. Opportunities checks those answers too, and the **AI Cited** column shows how often your domain appears in them.

That's the "AEO" half of SEO & AEO: answer engine optimization. A topic where competitors are cited and you aren't is a concrete gap.

## Setting Up

The first time you open the page, click **Set Up Research**. Research is organized into **projects**, each targeting one [ICP](/engage/content-engine/icp) or domain plus the competitors you want compared against.

Use **New Project** to research a different ICP or domain in its own project. Once you have more than one, a **Project** dropdown appears, including an **All Projects** view.

## Running Research

Click **Run Research**. It works through several phases — analyzing search results, checking AI assistant answers, gathering keyword volumes, comparing domain authority, and scoring opportunities.

It usually takes a few minutes and you can leave the page while it runs.

You can only run research on a single project at a time, so pick one from the dropdown if you're in **All Projects** view. The per-run credit cost is shown on the button before you commit.

## Reading the Results

| Column          | What it tells you                            |
| --------------- | -------------------------------------------- |
| **Opportunity** | The topic                                    |
| **Status**      | New, Generated, or Dismissed                 |
| **Type**        | What kind of content it calls for            |
| **Intent**      | What the searcher is trying to do            |
| **Stage**       | Which funnel stage it fits                   |
| **Volume/mo**   | Monthly search demand                        |
| **Score**       | How worthwhile it is, all factors considered |
| **AI Cited**    | How often you're cited in AI answers for it  |

Click any row to open the opportunity in full, including its brief.

Filter by **Status** to separate what's **New** from what you've already **Generated** or **Dismissed**. Dismissing the ones you'll never write keeps the list honest.

## Turning an Opportunity into Content

Open an opportunity and generate from its brief. The draft lands in [Content](/engage/content-engine/content), where you review, edit, and publish it like anything else — using your [brand voice](/engage/content-engine/brand-voice) and [strategy](/engage/content-engine/content-strategy) if you've set them.

## Tips

1. **Trust Score over Volume.** A high-volume topic you can't compete for is not an opportunity.
2. **Look for the citation gaps.** Topics where competitors are cited by AI and you aren't are the clearest openings.
3. **Dismiss aggressively.** A list you've curated is one you'll use; an untouched list of 200 isn't.
4. **Re-run periodically, not constantly.** The landscape moves over months, not days — and each run costs credits.

## Next Steps

* [Write the content](/engage/content-engine/content)
* [Research keywords directly](/engage/content-engine/seo-keyword-research)
* [Compare domain authority](/engage/content-engine/seo-domain-overview)


# Site Audit
Source: https://docs.parallellabs.app/engage/content-engine/seo-site-audit

Crawl your site and check every page for on-page SEO problems.

A site audit crawls your pages and reports what's wrong with them — missing titles, absent H1s, thin content, and other on-page issues that quietly cost you rankings.

## Running an Audit

1. Select your company from the sidebar
2. Go to **Content Engine** → **Site Audit**
3. Enter a **Domain**
4. Set **Max pages** — how far the crawl should go
5. Click **Run audit**

The crawl runs in the background and the page shows progress while it works. You can leave and come back. Cancelling an audit still keeps the findings for the pages it reached.

Use **Recent audits** to reload a previous run and compare.

## Reading the Results

**Issues found** lists each problem type with a severity and the number of affected pages. Click an issue to filter the pages table to just those pages, and **Clear filter** to go back.

The **Pages** table shows each crawled page's **URL**, **Title**, word count, **H1**, and issue count.

Work the issue list top-down by severity. One high-severity issue across forty pages is usually a template problem worth a single fix, not forty separate ones.

## Credits

An audit costs **25 credits as a base, plus 1 credit per page** crawled. **Max pages** is therefore your cost control, and it's capped at a per-run maximum shown in the form.

Crawling 100 pages costs 125 credits. If your site is large, audit the section you're working on rather than everything.

## Tips

1. **Start with a small crawl.** 25 pages usually surfaces the same template-level problems as 500.
2. **Fix templates, not pages.** Most repeated issues come from one layout.
3. **Re-audit after fixing** to confirm the change actually landed.
4. **Audit before a content push**, so new pages don't inherit an existing structural problem.

## Next Steps

* [Check your domain's authority](/engage/content-engine/seo-domain-overview)
* [Find topics worth writing about](/engage/content-engine/seo-opportunities)
* [Write the content](/engage/content-engine/content)


# Video
Source: https://docs.parallellabs.app/engage/content-engine/video

Generate short videos from text prompts.

Create videos using AI - from text descriptions or by animating images. Perfect for social media content, product showcases, marketing materials, and creative projects.

## What Can You Create?

* Social media video clips
* Product animations
* Background and ambiance videos
* Marketing and promotional content
* Animated scenes
* Creative visual content

## Creating Your First Video

### Step 1: Open Video Generation

1. Select your company from the sidebar
2. Click **Video** (under the Content Engine menu)
3. Click **Generate Video** (or the **+** button)

### Step 2: Choose Your Approach

You have two main options:

**Text-to-Video:** Describe what you want in words
**Image-to-Video:** Animate a starting image

### Step 3: For Text-to-Video

Write a detailed description of what you want to see:

```
A serene lake at sunset with gentle ripples on the water.
Mountains in the background with soft golden light.
A small boat floating peacefully. Cinematic quality.
```

### Step 4: For Image-to-Video

1. Click **Add Image**
2. Upload a starting image or select from your library
3. Describe how you want it to animate:

```
The clouds slowly moving across the sky,
leaves gently swaying in the wind.
```

### Step 5: Configure Settings

Choose your video options:

* **Duration:** Varies by model (typically 4–30 seconds)
* **Aspect ratio:** 16:9 (landscape), 9:16 (portrait/stories), 1:1 (square)
* **Resolution:** 720p or 1080p
* **Model:** Different AI models for different styles

### Step 6: Generate

1. Click **Generate**
2. Wait for processing (can take 1-5 minutes)
3. Preview your video when ready

## Understanding Video Settings

### Duration

Available durations depend on the model you pick. Common choices:

| Duration          | Best For                         |
| ----------------- | -------------------------------- |
| **5 seconds**     | Social clips, GIFs, quick cuts   |
| **10 seconds**    | Short stories, product showcases |
| **15-30 seconds** | Full scenes, advertisements      |

**Note:** Longer videos take more time to generate and use more credits.

### Aspect Ratio

| Ratio    | Dimensions | Best For                         |
| -------- | ---------- | -------------------------------- |
| **16:9** | Landscape  | YouTube, websites, presentations |
| **9:16** | Portrait   | TikTok, Instagram Stories, Reels |
| **1:1**  | Square     | Instagram feed, Facebook         |

### Resolution

| Resolution | Quality | File Size |
| ---------- | ------- | --------- |
| **720p**   | Good    | Smaller   |
| **1080p**  | High    | Larger    |

### Motion Settings

Some models offer additional controls:

| Setting              | What It Does                   |
| -------------------- | ------------------------------ |
| **Motion Intensity** | Low = subtle, High = dynamic   |
| **Camera Motion**    | Static, pan, zoom, orbit       |
| **Style**            | Realistic, cinematic, animated |

## Writing Great Video Prompts

### Be Specific About Movement

The AI needs to know what should move and how:

**Avoid — Vague:**
"A city scene"

**Better — Specific:**
"Cars driving slowly down a city street at night,
neon signs glowing, light rain falling,
camera slowly panning right"

### Include These Elements

| Element         | Example                           |
| --------------- | --------------------------------- |
| **Subject**     | "A woman in a red dress"          |
| **Action**      | "walking along the beach"         |
| **Environment** | "at sunset, waves gently rolling" |
| **Camera**      | "camera following from behind"    |
| **Style**       | "cinematic film quality"          |
| **Mood**        | "peaceful and romantic"           |

### Example Prompts

**Product showcase:**

```
A sleek smartphone rotating slowly on a white surface.
Soft studio lighting with subtle reflections.
Professional product video style, smooth motion.
```

**Nature scene:**

```
Aerial drone shot flying over a dense green forest.
Morning mist between the trees, golden sunlight
breaking through. Slow, smooth forward movement.
Cinematic quality.
```

**Abstract:**

```
Colorful abstract liquid shapes slowly morphing
and flowing. Deep blues and purples blending.
Smooth, mesmerizing motion. Perfect loop.
```

## Using Images to Start

### Why Start from an Image?

* More control over the visual style
* Use your own product photos
* Create consistent branded content
* Animate existing artwork or designs

### How to Use Image-to-Video

1. Upload or select your starting image
2. Describe what motion you want:
   ```
   The person's hair gently blowing in the wind,
   clouds moving slowly in the background
   ```
3. Configure settings
4. Generate

### First Frame / Last Frame

Some models let you specify both the start and end:

1. Upload a **first frame** (starting image)
2. Upload a **last frame** (ending image)
3. The AI creates a smooth transition between them

## Video Processing Time

Video generation takes longer than images or audio:

| Duration    | Typical Wait |
| ----------- | ------------ |
| 5 seconds   | 1-2 minutes  |
| 10 seconds  | 2-3 minutes  |
| 15+ seconds | 3-5+ minutes |

You can leave the page while waiting - check back later for your result.

### Checking Status

Look for status indicators:

* **Generating:** Still processing
* **Complete:** Ready to view
* **Failed:** Something went wrong (try again or adjust prompt)

## Managing Your Videos

### Viewing Your Library

1. Go to **Video**
2. Browse your generated videos
3. Click to preview any video

### Playing Videos

1. Click on a video
2. Use the video player controls
3. Full screen if needed

### Downloading Videos

1. Click on a video
2. Click **Download**
3. Video saves to your computer as MP4

### Deleting Videos

1. Click the **three-dot menu** (⋮) on the video
2. Select **Delete**
3. Confirm

## Tips for Better Videos

### 1. Start Short

Generate 5-second clips first to test your prompt. Once you like the result, try longer versions.

### 2. Keep Movement Simple

AI handles simple, single-direction movement best:

* Slow pan across a scene
* Gradual zoom in
* Single subject moving

Complex multi-element movement often has issues.

### 3. Be Patient

Video generation is computationally intensive. Give it time, and don't generate the same thing multiple times simultaneously.

### 4. Iterate on Your Prompt

If results aren't right:

* Add more specific details
* Simplify what you're asking for
* Try different style keywords
* Use a different model

### 5. Consider Combining Clips

For longer videos:

1. Generate multiple short clips
2. Download them
3. Combine in a video editor

## Common Issues

### "My video looks different than expected"

* Add more descriptive detail
* Specify the style explicitly
* Try a different model
* Reduce the number of moving elements

### "Motion is too fast/slow"

* Adjust the motion intensity setting
* Add "slow" or "gentle" to your prompt
* Shorter videos often have smoother motion

### "Video generation failed"

* Try a simpler prompt
* Reduce video length
* Check that your prompt doesn't violate content policies
* Try again later (might be high demand)

## Frequently Asked Questions

**Q: Can I add audio to generated videos?**
A: Generate video and [audio](/engage/content-engine/audio) separately, then combine in a video editor.

**Q: Why is video generation slow?**
A: Video requires generating many frames and is computationally intensive. It's normal for it to take a few minutes.

**Q: Can I use these videos commercially?**
A: Yes, check your subscription terms for specific commercial use rights.

**Q: What's the maximum video length?**
A: Typically 30 seconds per generation. For longer videos, generate multiple clips and combine them.

**Q: Can I generate videos with specific people?**
A: Generating realistic videos of specific real people isn't supported for privacy and ethical reasons.

## Next Steps

* [Create Audio to Add to Videos](/engage/content-engine/audio)
* [Generate Images](/engage/content-engine/images)
* [Clone Custom Voices](/engage/content-engine/voices)


# Voices
Source: https://docs.parallellabs.app/engage/content-engine/voices

Clone custom voices or use ready-made ones for your audio.

Clone voices to create unique, personalized text-to-speech that sounds like you, a team member, or a custom brand voice. Use these voices for consistent audio content across all your projects.

> **You don't have to clone a voice to get started.** A large library of ready-made voices is available to use immediately in [Audio](/engage/content-engine/audio). Clone a voice only when you want a specific person's or brand's voice. Cloned voices are private to your company; ready-made voices are available to everyone.

## What Is Voice Cloning?

Voice cloning uses AI to learn from audio samples and create a synthetic voice that sounds like the original speaker. Once created, you can use this voice to generate any text as speech.

### Why Clone a Voice?

* **Brand consistency:** One voice across all content
* **Personalization:** Use your own voice without recording everything
* **Scalability:** Generate hours of content from minutes of samples
* **Multiple languages:** Some voices can speak languages the original speaker doesn't

## Creating a Voice Clone

### Step 1: Open Voice Cloning

1. Select your company from the sidebar
2. Click **Voices** (under the Content Engine menu)
3. Click **Create Voice** (or the **+** button)

### Step 2: Prepare Your Audio Samples

You'll need to provide audio samples for the AI to learn from.

**Requirements:**

* High-quality recordings
* Clear speech without background noise
* 1-5 minutes of audio total
* Natural conversational speech

### Step 3: Upload Your Samples

1. Click **Upload Audio**
2. Select your audio file(s)
3. Supported formats: MP3, WAV, M4A

### Step 4: Configure Settings

* **Voice Name:** What you'll call this voice
* **Description:** Notes about the voice (optional)
* **Remove Background Noise:** Enable if recordings have some noise

### Step 5: Create the Voice

1. Click **Create Voice**
2. Wait for processing (can take several minutes)
3. Your voice is ready when the status shows complete

## Recording Great Voice Samples

The quality of your clone depends on the quality of your samples.

### Recording Environment

| Factor         | Recommendation                       |
| -------------- | ------------------------------------ |
| **Room**       | Quiet, no echo (closets work great!) |
| **Background** | No music, AC noise, or traffic       |
| **Time**       | Record when the environment is quiet |

### Recording Equipment

**Good options:**

* Professional microphone
* Quality headset microphone
* Modern smartphone in a quiet room

**Avoid:**

* Built-in laptop microphones
* Speakerphone recordings
* Phone calls or video conference recordings

### What to Say

Record natural, conversational speech:

* Read a few paragraphs from a book or article
* Tell a story about your work
* Explain something you know well
* Answer questions out loud

**Include variety:**

* Statements and questions
* Different emotions (neutral, happy, serious)
* Various sentence lengths

**Avoid:**

* Whispering or shouting
* Extreme emotions
* Heavy accents you don't normally use
* Reading in a monotone voice

### Sample Script

If you need something to read, try variations like:

```
"Hello and welcome! I'm excited to share some information
with you today. Did you know that our platform can help
you work more efficiently? Let me explain how that works.

First, you'll want to understand the basics. The system
is designed to be intuitive and easy to use. Whether
you're new to this or have years of experience, you'll
find value here.

I'd love to hear what questions you have. Feel free to
reach out anytime. Thanks for listening!"
```

## Testing Your Voice Clone

After creation, test before using in production:

### Step 1: Generate Test Audio

1. Find your voice in the list
2. Click **Test** or the play button
3. Enter test text:

```
This is a test of my cloned voice. Let me hear how
natural it sounds when reading this sentence.
```

4. Click **Generate** and listen

### Step 2: Evaluate the Result

Check for:

* Does it sound like the original speaker?
* Is pronunciation correct?
* Does it sound natural, not robotic?
* Are there any strange artifacts?

### Step 3: Refine If Needed

If quality isn't great:

* Try uploading more sample audio
* Ensure samples are high quality
* Remove any samples with background noise
* Re-create the voice with better samples

## Using Your Custom Voice

Once your voice is created, use it anywhere you generate audio:

### In the Audio Generator

1. Go to **Audio**
2. Click **Generate Audio**
3. In the voice selection, find your custom voice
4. Write your text and generate

### Your voice works for:

* All text-to-speech generation
* Any length content
* Multiple languages (if the model supports it)

## Managing Your Voices

### Viewing Your Voices

1. Go to **Voices**
2. See all your created voice clones
3. Status shows if each is ready to use

### Deleting a Voice

1. Click the **three-dot menu** (⋮) on the voice
2. Select **Delete**
3. Confirm

**Note:** Deleting a voice is permanent. Audio already generated with it will still work.

## Voice Quality Tips

### 1. More Samples = Better Quality

While 1 minute can work, 3-5 minutes of varied speech usually produces better results.

### 2. Consistent Recording Conditions

If recording multiple sessions:

* Use the same microphone
* Same distance from mic
* Same room/environment
* Similar time of day (voice changes throughout the day)

### 3. Natural Speech Patterns

Don't over-enunciate or speak unnaturally. The AI learns from how you actually speak, so be natural.

### 4. Clean Audio Editing

If you edit your audio:

* Remove long silences
* Cut out coughs, "ums," and mistakes
* Don't over-process with effects

## Ethical Use Guidelines

### DO use voice cloning for:

* Creating your own voice clone
* Voices of people who have given consent
* Fictional or synthetic brand voices
* Accessibility applications

### DON'T use voice cloning for:

* Impersonating others without consent
* Creating deceptive content
* Fraud or scams
* Spreading misinformation

**Important:** Only clone voices of people who have given explicit permission.

## Frequently Asked Questions

**Q: How long does voice creation take?**
A: Usually 5-15 minutes for processing.

**Q: Can I update a voice with more samples?**
A: Currently, you'd need to create a new voice with the combined samples.

**Q: How realistic are voice clones?**
A: Quality varies based on your samples. Good samples produce very realistic results.

**Q: Can the cloned voice speak other languages?**
A: Some models support multiple languages, even if your samples are in one language.

**Q: Is there a limit to how many voices I can create?**
A: Check your subscription for limits on voice clones.

**Q: Can I share my voice clone with others?**
A: Voice clones are currently limited to your account/company.

## Next Steps

* [Generate Audio with Your Voice](/engage/content-engine/audio)
* [Create Videos](/engage/content-engine/video)
* [Generate Images](/engage/content-engine/images)


# Inbox
Source: https://docs.parallellabs.app/engage/inbox

Connect sending mailboxes, warm them up, and manage replies with automatic classification.

The Inbox feature lets you connect and manage email accounts for sending sequences and campaigns. You can add multiple inboxes, set up custom domains, and warm up new accounts to improve deliverability.

## Why Do You Need an Inbox?

To send emails through sequences, you need to connect an email account. The Inbox feature:

* Connects your email accounts
* Manages sending limits
* Tracks email performance
* Warms up new accounts for better deliverability

## Connecting Your First Inbox

### Step 1: Add an Inbox

1. Select your company from the sidebar
2. Click **Inbox**
3. Click **Add Inbox** (or the **+** button)
4. Choose your email provider:
   * **Gmail/Google Workspace**
   * **Outlook/Microsoft 365**
   * **Other (SMTP/IMAP)**

### Step 2: Authorize Access

For Gmail or Outlook:

1. Click to connect
2. Sign in to your email account
3. Grant permission for the platform to send on your behalf
4. You're connected!

For SMTP/IMAP:

1. Enter your email address
2. Enter SMTP server details (your email provider can give you these)
3. Enter your password or app password
4. Test the connection
5. Save

### Step 3: Configure Settings

After connecting:

* Set your daily sending limit (start low, increase over time)
* Add your email signature
* Configure reply-to address if different

## Understanding Your Inbox Dashboard

When you view your inboxes, you'll see:

| Column      | What It Shows                     |
| ----------- | --------------------------------- |
| **Email**   | The connected email address       |
| **Status**  | Connected, Disconnected, or Error |
| **Sent**    | Emails sent recently              |
| **Health**  | Deliverability score              |
| **Warming** | Warming status (if enabled)       |

### Status Indicators

* **Green checkmark:** Connected and working
* **Yellow warning:** Needs attention
* **Red error:** Disconnected or problem

## Email Warming

New email accounts or domains need to build reputation before sending at scale. Email warming sends small amounts of email gradually to establish trust.

### Why Warm Up?

If you send too many emails from a new account:

* Emails may go to spam
* Your account could get flagged
* Deliverability suffers

Warming prevents these issues.

### Enabling Warming

1. Click on your inbox
2. Find **Warming** settings
3. Turn warming **On**
4. Configure:
   * **Daily limit:** Start with 5-10 emails/day
   * **Ramp rate:** How quickly to increase
5. Save

### How Warming Works

1. The system sends emails to other warming accounts
2. Those accounts open, read, and sometimes reply
3. This builds positive engagement signals
4. Your sender reputation improves
5. Gradually increase your limits

### Warming Timeline

| Week | Recommended Daily Volume |
| ---- | ------------------------ |
| 1-2  | 5-10 emails              |
| 3-4  | 15-25 emails             |
| 5-6  | 30-50 emails             |
| 7+   | 50-100+ emails           |

**Note:** These are guidelines. Monitor your deliverability and adjust.

## Setting Up a Custom Domain

Using your own domain ([you@yourcompany.com](mailto:you@yourcompany.com)) looks more professional and builds your brand reputation.

### Step 1: Add Your Domain

1. Click **Domains** in the Inbox section
2. Click **Add Domain**
3. Enter your domain name (e.g., yourcompany.com)
4. Click **Add**

### Step 2: Configure DNS Records

You'll need to add DNS records to verify ownership and improve deliverability:

**SPF Record:** Authorizes the platform to send on your behalf

```
Type: TXT
Host: @
Value: (provided in the setup wizard)
```

**DKIM Record:** Adds a digital signature to your emails

```
Type: TXT
Host: (provided, usually something._domainkey)
Value: (provided in the setup wizard)
```

**DMARC Record:** Tells receiving servers how to handle failed authentication

```
Type: TXT
Host: _dmarc
Value: v=DMARC1; p=none; (or as provided)
```

### Step 3: Add Records to Your DNS

1. Log into your domain registrar (GoDaddy, Namecheap, Cloudflare, etc.)
2. Find DNS settings
3. Add each record as specified
4. Save changes

### Step 4: Verify

1. Return to the platform
2. Click **Verify Domain**
3. Wait for verification (can take up to 48 hours, usually faster)
4. Green checkmark = verified!

## Viewing and Managing Messages

### Checking Your Inbox

1. Click on an inbox to view messages
2. Use filters to see:
   * **Received:** Incoming emails
   * **Sent:** Outgoing emails
   * **Bounced:** Failed deliveries
   * **Replies:** Responses to your sequences

### Reading a Message

1. Click on any message in the list
2. The full content appears in a preview
3. See the full conversation thread

### Composing a New Email

1. Click **Compose** or the compose button
2. Enter the recipient
3. Write your subject and message
4. Click **Send**

### Replying to Messages

1. Open a received message
2. Click **Reply**
3. Write your response
4. Click **Send**

## Managing Multiple Inboxes

You can connect multiple email accounts:

### When to Use Multiple Inboxes

* **Different team members:** Each person has their own inbox
* **Different purposes:** Sales vs. support vs. marketing
* **Higher volume:** Spread sending across accounts
* **A/B testing:** Test which sender gets better results

### Best Practices

1. **Don't over-send from one account** - Stay under provider limits
2. **Use consistent domains** - Build reputation for your domain
3. **Monitor each inbox** - Watch for issues in any account
4. **Rotate if needed** - Spread campaigns across inboxes

## Troubleshooting

### "Inbox disconnected"

1. Click on the inbox
2. Click **Reconnect** or **Reauthorize**
3. Sign in again to your email
4. Grant permissions

This happens when passwords change or permissions expire.

### "Emails going to spam"

* Enable and continue warming
* Check DNS records are correct
* Reduce sending volume
* Improve email content (less spammy language)
* Remove unengaged contacts from lists

### "High bounce rate"

* Clean your contact lists
* Verify email addresses before importing
* Remove invalid addresses
* Check for typos in email addresses

### "Not receiving replies"

* Check spam/junk folder
* Verify reply-to address is correct
* Make sure inbox is connected and syncing

## Daily Sending Limits

Email providers have limits on how many emails you can send:

| Provider         | Typical Daily Limit |
| ---------------- | ------------------- |
| Gmail (free)     | 500/day             |
| Google Workspace | 2,000/day           |
| Outlook (free)   | 300/day             |
| Microsoft 365    | 10,000/day          |

**Our recommendation:** Stay well under these limits. Send 50-100/day per inbox for best deliverability.

## Tips for Better Email Deliverability

### 1. Warm Up New Accounts

Always warm up new inboxes for 2-4 weeks before campaigns.

### 2. Maintain Good Hygiene

* Remove bounced addresses immediately
* Clean inactive contacts regularly
* Honor unsubscribe requests

### 3. Write Quality Emails

* Avoid spam trigger words
* Don't use all caps or excessive punctuation
* Include a clear unsubscribe option
* Balance text and images

### 4. Monitor Your Metrics

Watch open rates and bounce rates. If they drop suddenly, investigate.

### 5. Use Authentication

Set up SPF, DKIM, and DMARC for your domain.

## Frequently Asked Questions

**Q: How many inboxes do I need?**
A: Start with one. Add more as your volume grows or you need different senders.

**Q: Can I use my personal Gmail?**
A: Yes, but a professional email ([you@company.com](mailto:you@company.com)) looks more credible.

**Q: What happens if I exceed sending limits?**
A: Your email provider may temporarily block sending. Emails in sequences will queue until limits reset.

**Q: How long does domain verification take?**
A: Usually 15 minutes to 24 hours. Sometimes up to 48 hours for DNS propagation.

## Next Steps

* [Create Sequences](/engage/sequences)
* [Build Smart Lists](/find-enrich/smart-lists)
* [View Your Dashboard](/overview/dashboard)


# Keyword Monitoring
Source: https://docs.parallellabs.app/engage/keyword-monitoring

Track keywords across the web and social platforms, and get alerted when someone mentions them.

Track keywords, product names, and brand mentions across the web and social platforms. Get notified when something relevant is posted, see the sentiment, and generate a suggested reply in seconds.

## What Is Keyword Monitoring?

You create **monitors** that watch for your keywords across platforms like Reddit, Hacker News, LinkedIn, X, Quora, and AI assistants. When a matching post appears, it shows up as a **mention** with engagement metrics, a sentiment read, and an AI relevance score. You can generate reply suggestions on the spot, get notified by email or in-app, and optionally sync new mentions into a Smart List.

The section has two pages:

* **Monitors** — the watchers you set up: what they track, where, and how often.
* **Mentions** — what they found, and what you do about it.

Both pages share a stats row across the top — **New Mentions**, **Seen**, **Dismissed**, and a count per platform — plus a **Mentions by platform** chart over time. It's the quickest way to see whether a monitor is producing signal or noise before you read a single post.

## Monitors

### Creating a Monitor

1. Select your company from the sidebar
2. Click **Keyword Monitoring** (under Engage)
3. On the **Monitors** tab, click **New Monitor**
4. Fill in the details:
   * **Monitor Name** — e.g., "Brand mentions"
   * **Keywords** — add them comma-separated (press Enter or click **Add**)
   * **Platforms** — pick one or more: Reddit, Hacker News, LinkedIn, X, Quora, ChatGPT, Google AI
   * **Check Frequency** — Hourly, every 3 / 6 / 12 hours, Daily, or Weekly
5. *(Optional)* Refine it:
   * **Subreddit filter** (Reddit only) — limit to specific subreddits, or leave blank for all of Reddit
   * **AI Filter** (on by default) — drops irrelevant mentions; add plain-language instructions like "Skip job postings"
   * **AI Prompt Template** (for ChatGPT / Google AI) — customize the question asked, using `{keyword}` as a placeholder
6. Choose **Notifications** — Email and/or In-App — and pick which team members to notify
7. *(Optional)* **Sync to Smart List** — new mentions are added as rows automatically, including suggested replies
8. Review the per-run **credit cost** shown on the form, then click **Create Monitor**

### Managing Monitors

On the **Monitors** tab, each monitor's menu (⋮) lets you:

* **Run Now** — trigger a check immediately, outside the schedule
* **Pause / Resume** — stop or restart monitoring
* **Edit** — change keywords, platforms, filters, or notifications
* **Delete** — remove the monitor permanently

Each card shows its status, platforms, keywords, total mentions found, and when it last ran.

## Mentions

1. Open the **Mentions** tab
2. Filter by monitor, status (New, Seen, Dismissed), or platform
3. Each mention card shows the post, matched keywords, sentiment, and engagement (upvotes, comments, author, time)
4. Click a mention to open it, then:
   * **Open Original** to view the post on its platform
   * **Dismiss** to hide it from your feed
   * **Generate Replies** — add a bit of context about your product, and get reply drafts in several styles (value-first, educational, conversational, problem-solver); copy the one you like or regenerate

## Credits

Most platforms (Reddit, Hacker News, LinkedIn, X, Quora) are included. A couple of AI-assistant platforms cost more per run because they query large models:

* **ChatGPT** — 6 credits per keyword, per run
* **Google AI** — 2 credits per keyword, per run

The form shows the total cost per run (keywords × platforms) before you save, so there are no surprises. Checking more often uses more credits — pick a frequency that matches how quickly you need to know.

## Tips

* **Start specific.** Broad keywords catch a lot of noise; product names and branded phrases surface the mentions that matter.
* **Lean on the AI filter.** Add an instruction describing what you *don't* want to see to cut irrelevant hits.
* **Sync to a list** when you want to follow up systematically — it turns mentions into a working pipeline.

## Next Steps

* [Build a Smart List](/find-enrich/smart-lists)
* [Create an AI Employee](/overview/employees)
* [Set Up Notifications in Collaborators](/settings/collaborators)


# Marketing Campaigns
Source: https://docs.parallellabs.app/engage/marketing-campaigns

Send a one-off broadcast email to a Smart List, using a saved template.

Send a single, tracked email to everyone on a Smart List — a newsletter, an announcement, or a one-time promotion — with built-in open and click tracking.

## What Is a Marketing Campaign?

A marketing campaign is a **one-time broadcast email** sent from a sender profile to the contacts on one Smart List. It's the right tool when you want to send *one* message to *many* people at once.

**How it differs from other features:**

* **Marketing Campaign** — one email, sent to a whole list, once. Best for newsletters and announcements.
* **[Sequence](/engage/sequences)** — a multi-step, per-person workflow (email, LinkedIn, SMS, and more) with delays between steps. Best for outreach that follows up over time.
* **[Content Engine](/engage/content-engine)** — creates the content itself; it doesn't send bulk email.

## Creating a Campaign

1. Select your company from the sidebar
2. Click **Marketing Campaigns** (under Engage)
3. On the **Campaigns** tab, click **New Campaign**
4. Fill in the details:
   * **Campaign Name** — internal label; recipients never see it
   * **Subject** — the email subject line
   * **Sender Profile** — who the email comes from (required)
   * **From Name** and **Reply-To** — optional overrides
   * **Smart List** — the audience for this campaign
   * **Advanced → Email field** — pick which list column holds the email address (auto-detected in most cases)
5. Build the email body:
   * **Visual Builder** — a drag-and-drop email editor
   * or the **HTML editor** for a WYSIWYG experience
6. Choose when to send:
   * **Send Now** — begins sending immediately
   * **Schedule for Later** — pick a date, time, and timezone
   * *(Optional)* set per-day **send windows** so email only goes out during chosen hours
7. *(Recommended)* Send yourself a **test** first to check formatting and links
8. Click **Save**, **Save & Schedule**, or **Save & Send Now**

> **Sender tip:** For bulk sending, use a dedicated sending profile (such as one backed by a high-volume email service). Personal inbox profiles work but show a warning — they aren't built for large sends.

## Using Templates

Reuse a proven layout instead of starting from scratch:

1. Open **Templates** and click **New Template**
2. Build the template like a campaign, putting `{{content}}` where the per-campaign message goes
3. To use it, choose **Use template** from that template's menu — the campaign form opens with its styling and defaults applied

See [Templates](/engage/marketing-campaigns/campaign-templates) for the full details.

## Campaign Statuses

| Status        | What It Means                                        |
| ------------- | ---------------------------------------------------- |
| **Draft**     | Created but not sending — edit freely                |
| **Scheduled** | Queued for a future send                             |
| **Sending**   | Actively going out (paced automatically)             |
| **Sent**      | Completed — recipients and metrics are locked        |
| **Failed**    | A send error occurred — you can fix and retry        |
| **Cancelled** | A scheduled send was cancelled — edit and reschedule |

Sends are paced automatically so your campaign goes out at a steady rate rather than all at once. Recipients are captured from the list when sending begins, so later changes to the list don't affect a campaign already in flight.

## Tracking Performance

Open a campaign to see its results, and use the metrics bar for an at-a-glance view across all campaigns:

| Metric      | What It Tells You                                      |
| ----------- | ------------------------------------------------------ |
| **Sent**    | Emails successfully delivered                          |
| **Opened**  | Recipients who opened, with open rate                  |
| **Clicked** | Recipients who clicked a link, with click rate         |
| **Failed**  | Delivery errors                                        |
| **Skipped** | Rows with no email, duplicates, or suppressed contacts |

The campaign detail view also includes a per-recipient table (search by email) and a timeline of sends, opens, and clicks.

## How It Connects

* **[Smart Lists](/find-enrich/smart-lists)** — every campaign targets one list; build and enrich it first
* **[Enrichment](/find-enrich/lead-data/enrichment)** — fill in missing emails before you send
* **[Integrations](/settings/integrations)** — sending profiles are backed by your connected email services

## Next Steps

* [Build a Smart List](/find-enrich/smart-lists)
* [Run a Sequence instead](/engage/sequences)
* [Connect an Email Integration](/settings/integrations)


# Templates
Source: https://docs.parallellabs.app/engage/marketing-campaigns/campaign-templates

Save a reusable email style and settings so every campaign doesn't start from scratch.

A template stores the parts of a [marketing campaign](/engage/marketing-campaigns) that stay the same every time — the layout, the styling, the default list and sender — leaving only the actual message to write.

## The `{{content}}` Placeholder

A template is a complete email with one hole in it. Put `{{content}}` where the per-campaign message should go, and everything around it — header, footer, styling, unsubscribe — is reused on every send.

That's the whole idea: design the wrapper once, write only the middle each time.

## Creating a Template

1. Select your company from the sidebar
2. Go to **Marketing Campaigns** → **Templates**
3. Click **New Template**
4. Fill in:
   * **Name** and **Description** — how you'll recognize it later
   * **Subject** — a default subject line
   * **Default List** — the Smart List campaigns usually go to
   * **Default Sender** — the sender profile to send from
   * The email body, with `{{content}}` where the message belongs
5. Save

The defaults are starting points, not locks — you can change the list or sender on any individual campaign.

## Using a Template

From a template's menu, choose **Use template**. The campaign form opens with the template's styling and defaults already applied, ready for you to write the message and send.

You can also **Edit** or **Delete** a template from the same menu.

## Tips

1. **Build one good template first.** A single well-designed wrapper used consistently beats five half-finished ones.
2. **Put everything recurring in the template** — logo, footer, social links, unsubscribe. If you retype it each campaign, it belongs here.
3. **Set the defaults you actually use.** Filling in the usual list and sender removes two decisions per send.
4. **Send yourself a test** after editing a template, before it goes to a real list.

## Next Steps

* [Send a campaign](/engage/marketing-campaigns)
* [Build the list to send to](/find-enrich/smart-lists)
* [Connect a sending mailbox](/engage/inbox)


# Choosing a Channel
Source: https://docs.parallellabs.app/engage/overview

Reach the people you've found — by email, across social and content channels, and by listening for the moments worth replying to.

Once you have contacts, this section is how you actually talk to them: one-to-one outreach, broadcast email, published content, and monitoring for mentions worth responding to.

## In This Section

* **[Sequences](/engage/sequences)** — Multi-step outreach across email, LinkedIn, SMS, and voice, with A/B testing and reply tracking.
* **[Inbox](/engage/inbox)** — Connect sending mailboxes, warm them up, and manage replies with automatic classification.
* **[Content Engine](/engage/content-engine)** — Plan a strategy, generate on-brand content, research keywords, and produce images, audio, and video.
* **[Marketing Campaigns](/engage/marketing-campaigns)** — Send a one-off broadcast email to a Smart List using a saved template.
* **[Keyword Monitoring](/engage/keyword-monitoring)** — Track keywords across the web and social platforms, and get alerted to mentions.

## Sequences vs. Marketing Campaigns

Both send email, but they're built for different jobs:

|                    | Sequences                           | Marketing Campaigns           |
| ------------------ | ----------------------------------- | ----------------------------- |
| **Shape**          | Multi-step, over days or weeks      | One send                      |
| **Stops on reply** | Yes                                 | No                            |
| **Channels**       | Email, LinkedIn, SMS, voice, agents | Email                         |
| **Best for**       | Prospecting and follow-up           | Announcements and newsletters |

Use a sequence when you want a conversation. Use a campaign when you want to tell everyone something once.


# Sequences
Source: https://docs.parallellabs.app/engage/sequences

Multi-step outreach across email, LinkedIn, SMS, and voice, with A/B testing and reply tracking.

Sequences are automated multi-step campaigns that help you reach out to prospects at scale. Set up a series of emails, LinkedIn messages, or other touchpoints, and the system sends them automatically over time.

## What Are Sequences?

A sequence is a series of messages sent to contacts over time. For example:

**Day 1:** Send introduction email
**Day 3:** Follow up if no reply
**Day 7:** Send case study
**Day 10:** Final follow-up

Instead of manually sending each message, you add people to the sequence and it runs automatically.

## Creating Your First Sequence

### Step 1: Start a New Sequence

1. Select your company from the sidebar
2. Click **Sequences**
3. Click **Create Sequence** (or the **+** button)
4. Choose how to create it:
   * **Manual:** Build it yourself step-by-step
   * **AI Generated:** Describe your goal and let AI create it

### Step 2: Name Your Sequence

Give it a clear name like:

* "New Lead Follow-up"
* "Demo Request - No Show"
* "Customer Onboarding"

### Step 3: Add Steps

1. Click **Add Step**
2. Choose a step type (see the full list below) — email, LinkedIn, SMS, voice, an AI agent, or an automation step
3. Configure the step (write your message, set timing)
4. Repeat to add more steps

### Step 4: Save Your Sequence

Click **Save** when you're done adding steps.

## Understanding Step Types

Sequences support many step types. Each step also has a **wait time** that controls how long after the previous step it runs.

### Messaging Steps

* **Email** — Send an email from your connected inbox (subject + body)
* **SMS** — Send a text message *(requires an SMS integration)*
* **Voice Message** — Drop a voicemail or send a voice message *(requires a voice integration)*

**Example email:**

```
Subject: Quick question about [Company]

Hi &#123;&#123;firstName&#125;&#125;,

I noticed you downloaded our guide on [topic].
I'd love to hear what you thought of it.

Do you have 15 minutes this week for a quick call?

Best,
[Your name]
```

### LinkedIn Steps

* **LinkedIn Message** — Send a direct message to a connection
* **LinkedIn Connection Request** — Send a connection request with a note
* **LinkedIn Profile Visit** — Visit their profile (they'll see you viewed them)

**Note:** LinkedIn steps require a connected LinkedIn account and have daily limits.

### AI Agent Steps

* **Agent Conversation** — Kick off an AI-driven conversation with the contact
* **Auto Agent Loop** — Let an autonomous agent handle follow-ups across channels for a set number of steps

### Automation & Integration Steps

* **Browser Task** — Run an automated web task per member *(requires a browser session)*
* **HTTP Request** — Call an external API or webhook
* **Add to List** — Add the member to a Smart List
* **Add to Sequence** — Move the member into another sequence
* **Add to Prosp / HeyReach / Salesforge Campaign** — Push the member to a connected outreach platform
* **Thanks.io Direct Mail** — Send a physical mail piece *(requires the Thanks.io integration)*
* **Add to Meta Ads Audience** — Add the member to a Meta Ads audience

> Steps that need an integration are shown with the integration they require. Connect it under [Integrations](/settings/integrations) first.

### Wait Steps

Every step has a wait time that controls the pacing between touches:

* **Wait 2 days** before the next email
* **Wait 1 week** for the final follow-up

Good timing matters — don't overwhelm people with too many messages too fast.

## Using Variables (Personalization)

Make your messages feel personal with variables:

| Variable           | What It Becomes                 |
| ------------------ | ------------------------------- |
| `{{firstName}}`    | The person's first name         |
| `{{lastName}}`     | Their last name                 |
| `{{company}}`      | Their company name              |
| `{{email}}`        | Their email address             |
| `{{custom_field}}` | Any custom field from your list |

**Example:**

```
Hi &#123;&#123;firstName&#125;&#125;,

I saw that &#123;&#123;company&#125;&#125; recently expanded into new markets.
Congratulations! I'd love to discuss how we can support your growth.
```

## Adding People to a Sequence

### From a Smart List

1. Go to your [Smart List](/find-enrich/smart-lists)
2. Select the contacts you want to add
3. Click **Actions** → **Add to Sequence**
4. Choose your sequence
5. Map the list columns to sequence variables
6. Confirm

### Importing Directly

1. Open your sequence
2. Click **Add Members** or **Import**
3. Upload a CSV file or paste data
4. Map columns to fields
5. Import

### One at a Time

1. Open your sequence
2. Click **Add Member**
3. Fill in their details
4. Add

## Managing Sequence Members

### Viewing Members

1. Open your sequence
2. Click **Members** or the members count
3. See everyone in the sequence and their progress

### Member Status

| Status        | What It Means                         |
| ------------- | ------------------------------------- |
| **Active**    | Currently moving through the sequence |
| **Completed** | Finished all steps                    |
| **Paused**    | Temporarily stopped                   |
| **Replied**   | They responded (usually auto-pauses)  |
| **Bounced**   | Email couldn't be delivered           |

### Pausing or Removing Someone

1. Find them in the members list
2. Click on their row or the action menu
3. Choose **Pause** or **Remove**

## Turning Sequences On and Off

### Enabling a Sequence

1. Open the sequence
2. Find the **Active** toggle
3. Turn it on

The sequence will now automatically send messages to members.

### Disabling a Sequence

1. Open the sequence
2. Turn off the **Active** toggle

All scheduled sends are paused. Turn it back on to resume.

## Tracking Performance

### Viewing Metrics

1. Open your sequence
2. Click **Metrics** or look for the analytics section

### Key Metrics

| Metric      | What It Tells You                            |
| ----------- | -------------------------------------------- |
| **Sent**    | Total messages sent                          |
| **Opened**  | How many opened your emails (with open rate) |
| **Clicked** | How many clicked a link (with click rate)    |
| **Replied** | How many responded                           |
| **Booked**  | Meetings booked via a scheduling link        |
| **Bounced** | Failed deliveries                            |

### Open Rate Benchmarks

* **Great:** 30%+
* **Good:** 20-30%
* **Needs work:** Under 20%

If rates are low, try:

* Better subject lines
* More personalization
* Different send times
* Cleaning your list

## A/B Testing

Test different versions of a step to see what performs best:

1. On an email step, enable **A/B Test**
2. Add two or more **variants** (for example, different subject lines or openers)
3. The sequence rotates the variants across members
4. Compare variant performance (opens, clicks, replies) in the metrics view

Use the winning variant to improve your other steps and future sequences.

## Approving Messages

If your sequence uses AI-personalized content, you can require review before anything sends:

1. Open the sequence and look for the **approvals** view (or turn on **Auto Approve** to skip review)
2. Review each pending message
3. Click **Approve** to send or **Reject** to cancel

**Why approvals?** To make sure AI-personalized messages are appropriate before they go out.

## Managing Replies

When someone replies to your sequence:

### Automatic Handling

By default, replies automatically pause that person in the sequence (so they don't get more automated messages).

### Viewing Replies

1. Go to **Sequences** or check your **Inbox**
2. Look for replied members
3. Read their responses and reply personally

### Reply Management

1. Click **Reply Management** in your sequence
2. See all replies in one place
3. Categorize as positive, negative, or neutral
4. Respond as needed

## Sender Profiles

Control who emails come from:

1. Click **Sender Profiles** in sequences
2. Add or select a profile:
   * **Name:** How you appear to recipients
   * **Email:** The sending address
   * **Signature:** Your email signature

**Tip:** Use a real person's name, not "Sales Team" - personal emails get more responses.

## Tips for Better Sequences

### 1. Keep It Short

3-5 steps is usually enough. More than that feels spammy.

### 2. Provide Value

Each message should offer something useful, not just ask for something.

### 3. Space Out Your Steps

Give people time to respond:

* First follow-up: 2-3 days
* Second follow-up: 4-5 days
* Final follow-up: 7+ days

### 4. Write Like a Human

Avoid corporate speak. Write like you're emailing a colleague.

### 5. Test Before Sending

Send test emails to yourself to check:

* Variables fill in correctly
* Links work
* Formatting looks good

## Frequently Asked Questions

**Q: How many people can I add to a sequence?**
A: This depends on your inbox limits and daily sending caps. Start with small batches.

**Q: What happens if someone unsubscribes?**
A: They're automatically removed from the sequence and won't receive more messages.

**Q: Can I edit a sequence while it's running?**
A: Yes, but changes only affect future sends, not messages already scheduled.

**Q: Why did someone stop receiving messages?**
A: They may have replied, bounced, unsubscribed, or completed all steps.

**Q: Can I run multiple sequences at once?**
A: Yes, but be careful not to add the same person to multiple sequences simultaneously.

## Next Steps

* [Set Up Your Inbox](/engage/inbox)
* [Build Smart Lists](/find-enrich/smart-lists)
* [View Your Dashboard](/overview/dashboard)


# Lead Data
Source: https://docs.parallellabs.app/find-enrich/lead-data

Look up and import contacts from external data sources.

**Lead Data** is where contacts come into the platform from outside it. Its four sub-pages cover three ways to find people and one place to see what they cost.

## In This Section

* **[Enrichment](/find-enrich/lead-data/enrichment)** — Look up one contact or company by any identifier you have and get back verified details.
* **[People Search](/find-enrich/lead-data/people-search)** — Build an audience from filters — industry, seniority, location, company size, and more.
* **[Sales Navigator](/find-enrich/lead-data/sales-navigator)** — Turn a LinkedIn Sales Navigator search into a list of leads.
* **[Credits](/find-enrich/lead-data/lead-credits)** — Your data credit balance and what each lookup costs.

## Which One Should I Use?

| You have                                            | Use                                                                           |
| --------------------------------------------------- | ----------------------------------------------------------------------------- |
| A person's email, phone, or LinkedIn URL            | [Enrichment](/find-enrich/lead-data/enrichment)                               |
| A company name or domain, and you want people there | [Enrichment](/find-enrich/lead-data/enrichment) — it previews matching people |
| A description of your ideal customer, but no names  | [People Search](/find-enrich/lead-data/people-search)                         |
| A Sales Navigator search you've already built       | [Sales Navigator](/find-enrich/lead-data/sales-navigator)                     |

All three save their results into a [Smart List](/find-enrich/smart-lists), so the next step is the same no matter which you use.

## About Data Credits

Enrichment, People Search, and Sales Navigator all draw on one shared pool of **data credits**. Credits never expire. See [Credits](/find-enrich/lead-data/lead-credits) for balances, costs, and how to top up.

If your plan includes lead data, there is nothing to buy and the Credits page won't appear in your menu.


# Enrichment
Source: https://docs.parallellabs.app/find-enrich/lead-data/enrichment

Look up a person or company and get back verified contact details.

Look up a person or company and get back verified contact details — emails, phone numbers, job titles, company data, and more — then save the results straight into a Smart List.

## What Is Enrichment?

The Enrichment page is a quick lookup tool. Give it whatever you know about a contact (an email, a phone number, a LinkedIn URL, a name, or a company), and it returns matching contact data you can act on.

It works two ways:

* **Exact match** — When you provide a strong identifier (email, phone, LinkedIn URL, or a full name plus company), you get back a single, fully enriched contact.
* **Company search** — When you provide just a company name or domain, you get a preview of people at that organization, plus an estimate of how many matches exist.

This is the fast, one-off companion to the enrichment **actions** in [Smart Lists](/find-enrich/smart-lists), which enrich many rows at once.

## Enriching a Contact

1. Select your company from the sidebar
2. Click **Enrichment** (under Find & Enrich)
3. Fill in one or more fields — you only need one:
   * **First Name / Last Name**
   * **Email** — the highest match rate
   * **Phone**
   * **Company Name**
   * **Company Domain** (e.g., `example.com`)
   * **LinkedIn URL**
4. *(Optional, for company searches)* Narrow the results:
   * **Seniority** — e.g., VP, Director, Manager
   * **Department** — e.g., Sales, Engineering
   * **Job Title** — type a title and press Enter to add it
5. Click **Enrich Contact**

A chip above the results tells you what you got: an **exact match**, or a **lead search** with an estimated count of matching people.

## Saving Results to a Smart List

1. In the results table, tick the checkbox next to each contact you want (or use **Select All**)
2. Click **Add to Smart List**
3. Choose the target list from the dropdown — any missing columns are created for you automatically
4. Click **Add to List**

From there, the contacts are ready for [Sequences](/engage/sequences), [Marketing Campaigns](/engage/marketing-campaigns), or any Smart List action.

Use **Clear** to reset the form and start a new lookup.

## Credits

Enrichment draws on **data credits**:

* **1 credit** per contact returned on an exact match
* **1 credit** per company search that returns results

Failed lookups and empty results are free. Your current balance appears at the top of the page (on plans where enrichment isn't already included), and you can top up when you run low. If you're out of credits, the lookup will prompt you to add more.

## Tips

* **Lead with an email** when you have one — it gives the most reliable match.
* **Domain, not full URL** — enter `example.com`, not `https://www.example.com/about`. The page cleans it up either way.
* **Use the narrowing filters** for company searches to focus on the right seniority and departments before saving.

## Next Steps

* [Build Smart Lists](/find-enrich/smart-lists)
* [Run a Sequence](/engage/sequences)
* [Send a Marketing Campaign](/engage/marketing-campaigns)


# Credits
Source: https://docs.parallellabs.app/find-enrich/lead-data/lead-credits

Your data credit balance, what each kind of lookup costs, and how to top up.

**Data credits** pay for contact data from outside the platform. [Enrichment](/find-enrich/lead-data/enrichment), [People Search](/find-enrich/lead-data/people-search), and [Sales Navigator](/find-enrich/lead-data/sales-navigator) all draw on the same shared pool, and credits never expire.

> **If your plan includes lead data**, there is nothing to buy — the page shows **Included** and the Credits item may not appear in your menu at all.

## Checking Your Balance

1. Select your company from the sidebar
2. Go to **Lead Data** → **Credits**

Your remaining balance is shown at the top. Click **Buy Data Credits** to top up.

## What Things Cost

| Action              | Cost                                             |
| ------------------- | ------------------------------------------------ |
| **Enrichment**      | 1 credit per contact returned                    |
| **People Search**   | 1 credit per search, plus 1 per contact imported |
| **Sales Navigator** | 1 credit per lead the search returns             |

A lookup that finds nobody doesn't cost anything — you're charged for results, not attempts.

## Keeping Costs Down

1. **Give enrichment strong identifiers.** An email or LinkedIn URL matches far more often than a name alone, so fewer credits are spent on misses.
2. **Preview before importing.** Previewing an audience in People Search is cheap; importing it is what scales.
3. **Import in stages.** Pull 50, check they're right, then pull the rest.
4. **Cap Sales Navigator searches.** The maximum-leads slider is a direct cap on what a search can cost.
5. **Reuse saved results.** Recent lookups and recent searches are free to reopen.

## Data Credits vs. Plan Credits

These are not the same thing. **Data credits** buy contact data from external sources. The credits that pay for AI work — chat, content generation, agent conversations — are part of your subscription and are shown on your [Profile](/account/profile).

## Next Steps

* [Enrich a contact](/find-enrich/lead-data/enrichment)
* [Search for an audience](/find-enrich/lead-data/people-search)
* [View your subscription](/account/profile)


# People Search
Source: https://docs.parallellabs.app/find-enrich/lead-data/people-search

Build an audience from filters — industry, seniority, location, company size — and import the matches into a Smart List.

[Enrichment](/find-enrich/lead-data/enrichment) answers "tell me about this person." **People Search** answers the opposite question: "who fits this description?" You describe an ideal customer with filters, preview who matches, and import as many as you want.

## Building a Search

1. Select your company from the sidebar
2. Go to **Lead Data** → **People Search**
3. Give the audience a name — for example, "Marketing Directors in SF Bay Area"
4. Open the filter groups on the left and set your criteria
5. Click **Preview** to see matching contacts
6. Choose an **Import Size** and click **Import**
7. Pick the Smart List the contacts should land in

You need at least one filter before previewing. The header shows how many contacts match and how many data credits you have.

## Filter Groups

Filters are grouped down the left side. You can combine as many as you like — each one narrows the audience.

| Group                                                | Examples                                                                                                                                       |
| ---------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------- |
| **Business**                                         | Industries, Departments, Seniority, Job Titles, Exclude Job Titles, Company Names, Company Domains, Employee Count, Company Revenue, SIC Codes |
| **Location**                                         | Where the person or company is based                                                                                                           |
| **Contact**                                          | What contact details a record must have                                                                                                        |
| **Intent**                                           | Buying-intent signals, by segment or keyword                                                                                                   |
| **Personal**, **Financial**, **Housing**, **Family** | Consumer attributes, for B2C audiences                                                                                                         |

### Job title matching

**Job Titles** and **Exclude Job Titles** work together, and **Job Title Match** controls how strictly titles are compared. Excluding titles is often the fastest way to clean up an audience — screening out "Assistant" or "Intern" usually does more than adding another positive filter.

### Intent filters

Intent narrows to people showing buying signals. You can pick from ready-made audience segments or supply your own keywords, and there's an **AI Generator** to suggest keywords if you're not sure where to start.

## Previewing and Importing

**Preview** shows a sample of matching contacts on the right, with the total match count in the header. Treat the count as the audience size and the sample as a spot check — if the sample looks wrong, tighten the filters before importing.

**Import Size** offers 25, 50, 75, 100, 250, 500, 1,000, 5,000, 10,000, or **All**. Start small. Importing 50 and reading them tells you more about filter quality than importing 5,000 you haven't looked at.

## Credits

People Search costs **1 credit per search, plus 1 credit per contact imported**. Previewing the same search repeatedly is inexpensive; importing is what spends credits at volume. See [Credits](/find-enrich/lead-data/lead-credits).

## Tips

1. **Name your audiences descriptively.** "Mktg Dirs — Bay Area — 50-200 employees" beats "Test 3."
2. **Filter down, not up.** Start broad, preview, then add filters until the sample looks right.
3. **Import in stages.** Run outreach against a small import first; refine the filters before pulling the rest.
4. **Watch the match count.** A count in the millions means your filters aren't doing anything yet.

## Next Steps

* [Enrich a single contact](/find-enrich/lead-data/enrichment)
* [Search Sales Navigator instead](/find-enrich/lead-data/sales-navigator)
* [Work with the imported list](/find-enrich/smart-lists)


# Sales Navigator
Source: https://docs.parallellabs.app/find-enrich/lead-data/sales-navigator

Turn a LinkedIn Sales Navigator search into a list of leads — no LinkedIn account or cookie required.

If you already know how to build a good search in Sales Navigator, this page turns that search into contacts in a [Smart List](/find-enrich/smart-lists). You don't need to connect a LinkedIn account, and nothing runs through your own login.

## Two Ways to Search

### Paste a search URL

Build the search in Sales Navigator as you normally would, then copy the URL from your browser and paste it into **Sales Navigator search URL**.

The URL must contain `linkedin.com/sales/` — a regular LinkedIn search or a company page won't work.

### Build a search

If you'd rather not leave the platform, use the **Build a search** tab and set the filters directly:

* **Job title**
* **Location**
* **Current company**
* **School**
* **Keywords**

Set at least one. Filters combine, so each one you add narrows the search to a smaller, higher-quality set of leads.

## Running the Search

1. Select your company from the sidebar
2. Go to **Lead Data** → **Sales Navigator**
3. Paste a URL or set your filters
4. Set **Maximum leads to find** — anywhere from 1 to 500
5. Click **Search**

Larger searches take a few minutes. The results are saved as soon as the search finishes, so you can leave the page and reopen it later from **Recent searches**.

## Saving Results

Select the leads you want and click **Add to Smart List**. You can pick an existing list or type a new list name and create it on the spot.

People already on the list are matched on their LinkedIn URL and enriched in place, so re-running a search doesn't create duplicate rows.

## Credits

Sales Navigator uses **1 data credit per lead the search returns**, whether or not you add it to a list. A search that matches nobody is free.

Because the cost follows the result count, the **Maximum leads to find** slider is also your budget control. See [Credits](/find-enrich/lead-data/lead-credits).

## Tips

1. **Cap the maximum while testing.** Run 10 leads to check the search is right before running 500.
2. **Narrow with filters, not volume.** A tight search of 50 good leads beats 500 loose ones you'll never contact.
3. **Reuse past searches.** **Recent searches** keeps your results — no need to pay for the same search twice.

## Next Steps

* [Work with the imported list](/find-enrich/smart-lists)
* [Build an audience from filters instead](/find-enrich/lead-data/people-search)
* [Start a sequence](/engage/sequences)


# Where Contacts Come From
Source: https://docs.parallellabs.app/find-enrich/overview

Build the contact data you'll reach out to: unified profiles, working lists, and verified details from external sources.

Everything in this section answers one question — *who are we talking to?* Start from a Smart List you build by hand, or pull contacts in from external data sources, then enrich them until you have enough to run outreach.

## In This Section

* **[People](/find-enrich/people)** — Unified profiles that merge the same person across Smart Lists, sequences, campaigns, conversations, and your inbox.
* **[Smart Lists](/find-enrich/smart-lists)** — Spreadsheet-style tables of contacts and data, with per-row AI actions.
* **[Lead Data](/find-enrich/lead-data)** — Look up and import contacts from external sources: Enrichment, People Search, Sales Navigator, and the credits that pay for them.

## Where Contacts Come From

| Route                                                       | Best for                                           |
| ----------------------------------------------------------- | -------------------------------------------------- |
| Type or paste into a [Smart List](/find-enrich/smart-lists) | Contacts you already have                          |
| [Enrichment](/find-enrich/lead-data/enrichment)             | Filling in details for one known person or company |
| [People Search](/find-enrich/lead-data/people-search)       | Building an audience from scratch using filters    |
| [Sales Navigator](/find-enrich/lead-data/sales-navigator)   | Turning a LinkedIn search into a list              |
| A CRM [integration](/settings/integrations)                 | Syncing contacts you keep elsewhere                |

All of them land in the same place: a Smart List you can enrich, score, and feed into a [sequence](/engage/sequences).


# People
Source: https://docs.parallellabs.app/find-enrich/people

One profile per human being, merged from every place they appear in your account.

The same person often shows up in several places — a row in a Smart List, a recipient in a sequence, someone who replied in your inbox, a visitor who chatted with an agent. **People** stitches those together into a single profile so you can see everything you know about someone in one view.

> **Availability:** People is currently limited to super users while the identity graph is being proven out. Profiles merge records across every feature, so a bad merge would be visible to the whole company.

## How Profiles Are Built

Profiles are assembled by a background job that runs nightly. It looks for **identifiers** shared between records and merges the records that match.

Identifiers come in two strengths:

| Identifier         | Strength | Why                                                                                   |
| ------------------ | -------- | ------------------------------------------------------------------------------------- |
| **Email**          | Strong   | Unique to one person                                                                  |
| **Phone**          | Strong   | Unique to one person                                                                  |
| **LinkedIn**       | Strong   | Unique to one person                                                                  |
| **Name + Company** | Weak     | Names collide — two people called Chris Taylor at the same firm are indistinguishable |

Strong identifiers merge records with confidence. A **Name + Company** match is treated as low confidence: the profile is flagged **Needs review**, and the touchpoints it contributed are marked `low confidence` so you can judge them yourself.

The graph is read-only — you can't merge or split profiles by hand. Corrections come from fixing the underlying records, which the next nightly run picks up.

## Browsing People

1. Select your company from the sidebar
2. Click **People** (under Find & Enrich)

The table shows **Name** (with job title beneath), **Email**, **Phone**, **Company**, **Appears in**, and **Last seen**. Above it, a summary line counts the people, their total touchpoints, and how many need review.

* **Search** matches name, email, phone, or company
* **Source** filters to people who appear in a given place: Smart List, Sequence, Campaign, Conversation, or Inbox
* **Appears in** shows one chip per source with a count, so you can see at a glance where someone came from
* A warning triangle beside a name means a weak match contributed to the profile

If a nightly run is still in progress, a banner tells you how many records have been scanned so far.

## Reading a Profile

Click any row to open the person's profile. It shows:

* **Identity** — name, job title, company, website, and the identifier chips that built the profile
* **First seen** and **Last seen** dates
* **Activity** — a timeline of what this person actually did
* **Touchpoints** — one panel per source, listing the records they appear in

### The Activity Timeline

The timeline records real events: **Opens**, **Clicks**, **Scheduling views**, **Bookings**, **Status changes**, and **Calls**. Use the filter chips to narrow it to the kind of event you care about.

Automated traffic — email scanners and image proxies that open a message without a human reading it — is hidden by default. Toggle **Automated** to show it; those rows carry an **Automated** chip so you never mistake a scanner for a prospect.

Only the most recent touchpoints are shown on a busy profile.

## What People Is Good For

* Checking whether you're already talking to someone before starting fresh
* Seeing every prior touch before a sales call
* Spotting the same person sitting in three different lists under two email addresses
* Understanding which channel a lead actually engages with

## Next Steps

* [Build a Smart List](/find-enrich/smart-lists)
* [Enrich a contact](/find-enrich/lead-data/enrichment)
* [Run a sequence](/engage/sequences)


# Smart Lists
Source: https://docs.parallellabs.app/find-enrich/smart-lists

Spreadsheet-style tables of contacts and data, with per-row AI actions.

Smart Lists are powerful spreadsheet-like tools for organizing your data. Use them to manage contacts, track leads, organize tasks, or store any structured information. Unlike regular spreadsheets, Smart Lists can connect to AI features and automations.

## What Are Smart Lists Good For?

* **Lead lists:** Track prospects and their information
* **Customer databases:** Store customer details
* **Contact management:** Organize people to reach out to
* **Task tracking:** Manage to-do items and projects
* **Data collection:** Gather and organize any information

## Creating Your First List

1. Select your company from the sidebar
2. Click **Smart Lists**
3. Click **Create List** (or the **+** button)
4. Give your list a name (e.g., "Sales Prospects Q1")
5. Click **Create**

Your empty list opens and you can start adding columns and data.

## Setting Up Columns

Columns define what information you'll track. Click on a column header to configure it.

### Adding a Column

1. Click the **+** at the end of the column headers
2. Choose a column type
3. Name your column
4. Configure any options (like dropdown choices)

### Column Types

| Type             | Best For                    |
| ---------------- | --------------------------- |
| **Text**         | Names, short descriptions   |
| **Long Text**    | Notes, longer content       |
| **Number**       | Amounts, quantities, scores |
| **Email**        | Email addresses             |
| **URL**          | Website links               |
| **Date**         | Deadlines, dates            |
| **Select**       | Single choice from options  |
| **Multi-Select** | Multiple choices            |
| **Checkbox**     | Yes/no, done/not done       |

### Editing a Column

1. Click on the column header
2. Select **Edit Column**
3. Change the name, type, or options
4. Save

### Deleting a Column

1. Click on the column header
2. Select **Delete Column**
3. Confirm (data in that column will be lost)

## Adding Data

### Adding a Single Row

1. Click **Add Row** or the **+** at the bottom of the list
2. Fill in the fields
3. Data saves automatically

### Importing Data

Have data in a spreadsheet? Import it:

1. Click **Import** or the upload icon
2. Select your file (CSV or Excel)
3. Map your file's columns to list columns
4. Review the preview
5. Click **Import**

**Tips for smooth imports:**

* Make sure your file has headers in the first row
* Clean up your data before importing
* Match column types (don't put text in number fields)

### Copying and Pasting

You can often copy data from a spreadsheet and paste directly into a Smart List:

1. Select cells in your spreadsheet
2. Copy (Ctrl+C or Cmd+C)
3. Select a cell in your Smart List
4. Paste (Ctrl+V or Cmd+V)

## Editing Data

### Inline Editing

Click on any cell to edit it. Changes save automatically.

### Bulk Editing

1. Select multiple rows using checkboxes
2. Click **Bulk Edit** or the edit icon
3. Choose what to change
4. Apply to all selected rows

## Finding and Filtering Data

### Searching

Use the search bar to find rows containing specific text:

1. Click the search icon or search field
2. Type your search term
3. Results filter in real-time

### Filtering

Show only rows matching certain criteria:

1. Click **Filter** or the filter icon
2. Choose the column to filter by
3. Set the condition (equals, contains, greater than, etc.)
4. Apply

**Example filters:**

* Status equals "Active"
* Email contains "@gmail"
* Score is greater than 50

### Sorting

Organize your data by any column:

1. Click on a column header
2. Choose **Sort A→Z** or **Sort Z→A**
3. Data reorders

## List Actions

Smart Lists do far more than store data. **Actions** run automatically on each row — enriching contacts, scoring leads, generating content, syncing to your CRM, and more. Some actions add a new value to a cell; others add a **lookup column** that stays up to date.

### Enrichment & Data

| Action                        | What It Does                                                                                             |
| ----------------------------- | -------------------------------------------------------------------------------------------------------- |
| **Apollo Enrichment**         | Enrich rows with Apollo data, including email and phone reveal                                           |
| **Findymail Enrichment**      | Find and verify email addresses                                                                          |
| **Better Contact Enrichment** | Enrich contacts with emails and phone numbers                                                            |
| **Audience Enrichment**       | Append firmographic and demographic fields *(availability depends on your plan)*                         |
| **Find Leads at Company**     | Find new leads at a company by seniority, department, or job title *(availability depends on your plan)* |
| **CRM Lookup**                | Check whether a row already exists in your connected CRM                                                 |

### AI Actions

| Action                             | What It Does                                                    |
| ---------------------------------- | --------------------------------------------------------------- |
| **AI Web Scraper: Research Agent** | Research each row on the web and return structured findings     |
| **AI Text Generator**              | Generate custom text per row (summaries, openers, descriptions) |
| **AI Lead Scoring**                | Score each lead 0–100 based on criteria you define              |
| **AI Image Generator**             | Generate an image for each row                                  |
| **AI Audio Generator**             | Generate audio for each row                                     |

### Automation & Integrations

| Action                | What It Does                                                                |
| --------------------- | --------------------------------------------------------------------------- |
| **Add to Sequence**   | Add selected rows to an outreach sequence                                   |
| **Add to Smart List** | Copy rows into another list                                                 |
| **Add Tags**          | Apply tags, optionally based on conditions                                  |
| **Sync to CRM**       | Push rows to HubSpot or GoHighLevel                                         |
| **Browser Task**      | Run an automated web task for each row *(requires a browser session)*       |
| **HTTP Request**      | Call any external API or webhook (Zapier, Make, your own endpoints) per row |

### Lookup Columns

Lookup columns pull in live status from elsewhere in the platform:

* **Sequence Status** — shows where each contact stands in your sequences
* **Agent Conversations** — shows agent chat activity tied to each row

> The exact set of actions you see depends on your plan and which integrations you've connected.

### Running an Action

1. Select the rows you want to act on (or select all)
2. Click **Actions** or the action button
3. Choose your action
4. Configure its settings
5. Run it — results fill into the relevant column as each row completes

## Sharing Your List

### With Team Members

1. Click **Share** or the sharing icon
2. Choose sharing level:
   * **View:** They can see but not edit
   * **Edit:** They can make changes
   * **Full:** They can manage settings
3. Save

### Exporting Data

To get your data out:

1. Click **Export**
2. Choose format (CSV, Excel)
3. Download the file

## Organizing Multiple Lists

### Renaming a List

1. Go back to the Smart Lists overview
2. Click the **three-dot menu** (⋮) on the list
3. Select **Rename**
4. Enter the new name
5. Save

### Deleting a List

1. Click the **three-dot menu** (⋮) on the list
2. Select **Delete**
3. Confirm

**Warning:** Deleting a list removes all its data permanently.

## Connecting Lists to Other Features

### Adding to Sequences

Use your list data for email campaigns:

1. Open your list
2. Select the rows you want to contact
3. Click **Actions** → **Add to Sequence**
4. Choose or create a sequence
5. Map your list columns to sequence variables (first name, email, etc.)
6. Confirm

### Enriching Before You Reach Out

A common flow is to enrich a list first, then engage:

1. Import or build your list of contacts
2. Run enrichment actions (Apollo, Findymail, Better Contact) to fill in emails and phone numbers
3. Score or research rows with AI actions
4. Add the qualified rows to a [Sequence](/engage/sequences) to start outreach

## Tips for Effective Lists

### 1. Plan Your Columns First

Think about what data you need before creating columns. It's easier to set up correctly than to reorganize later.

### 2. Use Consistent Data

Keep data format consistent:

* Dates: Always use the same format
* Names: Decide on "John Smith" vs "Smith, John"
* Status: Use dropdowns to ensure consistent options

### 3. Keep It Clean

Regularly review and clean your lists:

* Remove duplicates
* Update outdated information
* Archive old data

### 4. Use Select Fields for Categories

Instead of typing "Active" or "Inactive" in a text field, use a Select column. This prevents typos and makes filtering easier.

## Frequently Asked Questions

**Q: How many rows can a list have?**
A: Lists can hold thousands of rows. Very large lists may load more slowly.

**Q: Can I undo changes?**
A: Individual cell edits don't have undo. For major changes, export a backup first.

**Q: Can I move rows between lists?**
A: Yes — use the **Add to Smart List** action to copy selected rows into another list.

**Q: Who can see my lists?**
A: By default, only you. Use sharing settings to give others access.

## Next Steps

* [Create Email Sequences](/engage/sequences)
* [Manage Your Inbox](/engage/inbox)
* [Connect Integrations](/settings/integrations)


# Getting Started
Source: https://docs.parallellabs.app/getting-started

An AI workforce for your business: AI employees that chat, research, enrich contacts, run outreach, create content, and answer your customers.

Welcome! This guide will help you get started and make the most of every feature available to you.

## Getting Started

When you first log in, you'll create a company (or organization) to get started. Think of a company as your workspace where all your AI employees, content, contacts, and campaigns live.

### Your First Steps

1. **Create Your First Company** - [Learn how](/account/companies)
2. **Set Up Your AI Employees** - [Learn how](/overview/employees)
3. **Add Knowledge to Train Your AI** - [Learn how](/overview/knowledge-base)
4. **Start Chatting with Your AI** - [Learn how](/overview/chat)

## What Can You Do?

### Talk to AI Employees

Create AI employees with different personalities and expertise, then chat with them to get help with tasks, answer questions, or brainstorm ideas.

* [Chat with your AI](/overview/chat)
* [Create AI Employees](/overview/employees)
* [Give them work that runs on its own](/overview/scheduled)

### Build Your Knowledge Base

Upload documents, websites, and files to train your AI employees. The more relevant knowledge you add, the more accurate your AI becomes.

* [Knowledge Base Guide](/overview/knowledge-base)

### Find & Enrich Contacts

Build lists of contacts, enrich them with verified emails, phone numbers, and firmographic data, and run automated per-row actions.

* [Build Smart Lists](/find-enrich/smart-lists)
* [Look up and import leads](/find-enrich/lead-data)
* [See unified profiles](/find-enrich/people)

### Run Outreach Campaigns

Send personalized emails, LinkedIn touches, SMS, and more through multi-step sequences that nurture your leads and book meetings.

* [Create Sequences](/engage/sequences)
* [Manage Email Inboxes](/engage/inbox)
* [Send a one-off broadcast](/engage/marketing-campaigns)

### Create Content

Plan a strategy, research what people search for, then generate blog posts, social content, images, audio, and video — and publish on a schedule.

* [Content Engine](/engage/content-engine)
* [Find topics worth writing about](/engage/content-engine/seo-opportunities)
* [Generate Images](/engage/content-engine/images)
* [Design social graphics](/engage/content-engine/canvas)

### Answer Customers Automatically

Deploy agents that answer on your website, over the phone, and in messaging apps — capturing leads around the clock.

* [Create Agents](/inbound/agents)
* [Connect WhatsApp](/inbound/agents/whatsapp)

### Build Sites and Apps

Describe what you want and get a working site, landing page, or app — published to your own domain.

* [Build a Website](/inbound/websites)
* [Create Landing Pages](/inbound/landing-pages)
* [Build an App](/inbound/app-builder)

### Listen for Mentions

Track your brand and keywords across social platforms and AI assistants, and reply to what matters.

* [Keyword Monitoring](/engage/keyword-monitoring)

## Navigation Guide

After selecting a company, the sidebar groups features into collapsible sections. Every item below has a page in these docs.

**At the top**

| Item                                       | What It's For                        |
| ------------------------------------------ | ------------------------------------ |
| [Employees](/overview/employees)           | Create and manage your AI employees  |
| [Chat](/overview/chat)                     | Talk to your AI employees            |
| [Scheduled](/overview/scheduled)           | Work your employees run on their own |
| [Knowledge Base](/overview/knowledge-base) | Upload training documents            |
| [Dashboard](/overview/dashboard)           | See your analytics and metrics       |

**[Find & Enrich](/find-enrich/overview)**

| Item                                    | What It's For                                           |
| --------------------------------------- | ------------------------------------------------------- |
| [People](/find-enrich/people)           | Unified profiles merged across every feature            |
| [Smart Lists](/find-enrich/smart-lists) | Manage contact lists and run per-row actions            |
| [Lead Data](/find-enrich/lead-data)     | Enrichment, People Search, Sales Navigator, and Credits |

**[Engage](/engage/overview)**

| Item                                               | What It's For                              |
| -------------------------------------------------- | ------------------------------------------ |
| [Sequences](/engage/sequences)                     | Run multi-step outreach campaigns          |
| [Inbox](/engage/inbox)                             | Manage connected mailboxes and replies     |
| [Content Engine](/engage/content-engine)           | Strategy, content, SEO research, and media |
| [Marketing Campaigns](/engage/marketing-campaigns) | Send one-time broadcast emails to a list   |
| [Keyword Monitoring](/engage/keyword-monitoring)   | Track keywords and brand mentions          |

**[Inbound](/inbound/overview)**

| Item                                    | What It's For                                        |
| --------------------------------------- | ---------------------------------------------------- |
| [Agents](/inbound/agents)               | Chatbots for your website, phone, and messaging apps |
| [Websites](/inbound/websites)           | Full marketing sites with a built-in CMS             |
| [Landing Pages](/inbound/landing-pages) | Single pages for campaigns and lead capture          |
| [App Builder](/inbound/app-builder)     | Build and publish custom apps                        |

**[Apps](/apps)**

Your published and installed apps appear here, each opening in the main area.

**[Settings](/settings/overview)**

| Item                                     | What It's For                        |
| ---------------------------------------- | ------------------------------------ |
| [Integrations](/settings/integrations)   | Connect external tools and services  |
| [Collaborators](/settings/collaborators) | Invite team members and manage roles |

**[At the bottom](/account/overview)**

| Item                        | What It's For                            |
| --------------------------- | ---------------------------------------- |
| [Profile](/account/profile) | Your account, subscription, and security |
| [Support](/account/support) | Get help                                 |

> The exact items you see depend on your plan and which features are enabled for your account. Some items — such as **People** and **Admin** — are limited to super users, and **Websites** is in beta.

## Account & Settings

* [Manage Your Companies](/account/companies)
* [Your Profile & Subscription](/account/profile)
* [Get Support](/account/support)

## Need Help?

* Click **Support** in the sidebar to submit a support ticket
* Our team typically responds within one business day
* Include screenshots when reporting issues — it helps us help you faster

## Quick Tips

1. **Start small** — create one AI employee and test chatting before building complex campaigns
2. **Quality over quantity** — a few well-written knowledge base documents beat many low-quality ones
3. **Test your sequences** — always send yourself a test before launching a campaign
4. **Save your work** — most changes save automatically, but click Save when you see the button
5. **Use search** — most pages have search to help you find what you need quickly


# Agents
Source: https://docs.parallellabs.app/inbound/agents

Deploy chatbots on your website, by phone, and in messaging apps to answer questions and capture leads.

Build AI-powered chatbots that you can add to your website. Agents handle visitor questions, capture leads, and provide instant support 24/7.

> **Agents vs. AI Employees:** An [AI Employee](/overview/employees) is the assistant you configure and chat with inside the app. An **Agent** is a public-facing chatbot you embed on your website — each agent is *powered by* one of your AI employees but has its own system prompt, appearance, and lead-capture settings.

## What Are Agents?

Agents are customer-facing assistants powered by your AI employees. Unlike Chat — which is for you and your team — an agent talks to the outside world. They:

* Answer visitor questions automatically
* Use your knowledge base to give accurate responses
* Capture leads and contact information
* Provide instant support without human intervention
* Work around the clock

## Channels

One agent can answer on several channels at once. Each channel is a tab in the agent editor, and each is enabled independently.

| Channel      | What it does                                                               |
| ------------ | -------------------------------------------------------------------------- |
| **Widget**   | A chat bubble embedded on your website                                     |
| **Webpage**  | A standalone, shareable chat page — no embedding required                  |
| **SMS**      | Replies to text messages sent to a connected number                        |
| **WhatsApp** | Replies to WhatsApp messages — see [WhatsApp](/inbound/agents/whatsapp)    |
| **Voice**    | Answers phone calls with a spoken voice, and can place outbound test calls |
| **Email**    | Replies to messages sent to a connected mailbox                            |

The other tabs configure the agent rather than adding a channel:

* **General** — name, system prompt, knowledge base access, contact capture, and response delay
* **Integrations** *(Beta)* — let the agent use connected tools while it talks
* **Handoff** — hand a conversation to a human when the agent can't help
* **Notifications** — who gets told about new conversations and captured leads

### Voice agents

Voice agents answer a phone number on a connected telephony integration. You choose the voice and language, and can let the agent collect contact details during the call, send an SMS mid-call, or hand off to a person. Use **Test Call** on the agent row to have it call you so you can hear it before going live.

## Conversations

The **Conversations** item under Agents shows every conversation your agents have had, across all channels, in a two-panel reader: the list on the left, the full transcript on the right. Use it to check answer quality, find captured leads, and spot questions the agent handled badly.

## Creating Your First Agent

### Step 1: Open Agents

1. Select your company from the sidebar
2. Click **Agents** (under Inbound)
3. Click **Create Agent** (or the **+** button)

### Step 2: Basic Settings

Fill in the essentials:

* **Name:** What you'll call this agent internally
* **Avatar:** Upload an image for your agent
* **Employee:** Select which AI employee powers this agent

### Step 3: Choose a System Prompt Template

Select how your agent should behave:

**Support Agent:**

* Answers questions based on your training data
* Stays focused on helping customers
* Professional and helpful tone

**Sales Assistant:**

* Qualifies prospects with questions
* Collects contact information
* Guides visitors toward next steps

**Custom:**

* Write your own instructions from scratch
* Full control over behavior

### Step 4: Configure Responses

* **Fallback Response:** What to say when the agent doesn't know the answer
* **Welcome Message:** The first message visitors see

**Example Welcome Message:**

```
Hi! I'm here to help answer your questions.
What can I assist you with today?
```

### Step 5: Save and Test

1. Click **Save** or **Create**
2. Use the **Preview** to test your agent
3. Refine settings as needed

## Adding the Agent to Your Website

### Getting the Embed Code

1. Go to your agent in the list
2. Click the **three-dot menu** (⋮)
3. Select **Get Embed Code**
4. Copy the code snippet

### Installing on Your Website

Paste the code before the `</body>` tag on your website:

```html theme={null}
<script>
  window.agentConfig = &#123;
    agentId: 'your-agent-id',
    position: 'bottom-right',
    theme: 'light'
  &#125;;
</script>
<script src="https://yourplatform.com/widget.js"></script>
```

### Platform-Specific Instructions

**WordPress:**

1. Go to Appearance → Theme Editor
2. Edit footer.php
3. Paste the code before `</body>`

Or use a plugin like "Insert Headers and Footers"

**Shopify:**

1. Go to Online Store → Themes
2. Click Actions → Edit code
3. Find theme.liquid
4. Paste before `</body>`

**Wix:**

1. Go to Settings → Custom Code
2. Add new custom code
3. Paste the embed code
4. Set to load on all pages

**Squarespace:**

1. Go to Settings → Advanced → Code Injection
2. Paste in the Footer section
3. Save

**Other Platforms:**
Look for "Custom Code," "Footer Scripts," or "Code Injection" in your platform's settings.

## Customizing Your Agent's Appearance

### Widget Options

| Setting           | Options                   | Description                   |
| ----------------- | ------------------------- | ----------------------------- |
| **Position**      | bottom-right, bottom-left | Where the chat button appears |
| **Theme**         | light, dark, auto         | Color scheme                  |
| **Primary Color** | Any hex color             | Your brand color              |
| **Greeting**      | Custom text               | Initial message shown         |
| **Placeholder**   | Custom text               | Text in input field           |

### Updating Appearance

1. Edit your agent
2. Find the appearance settings
3. Customize colors, position, and messages
4. Save changes
5. Get new embed code if needed

## Connecting to Your Knowledge Base

Your agent can use your [Knowledge Base](/overview/knowledge-base) to answer questions accurately.

### How It Works

1. A visitor asks a question
2. The AI searches your knowledge base
3. Finds relevant information
4. Generates an accurate response based on your content

### Setting Up Knowledge Access

1. Make sure your AI employee has access to the knowledge base
2. Upload relevant documents to your knowledge base
3. The agent automatically uses this information

### Tips for Better Responses

* Add FAQ documents to your knowledge base
* Include product information and pricing
* Upload support articles and guides
* Keep information up to date

## Lead Capture

Collect visitor information to follow up later.

### Setting Up Lead Capture

1. Edit your agent
2. Find **Lead Capture** settings
3. Enable lead capture
4. Configure required fields

### Fields You Can Collect

* Name
* Email address
* Phone number
* Company name
* Custom fields

### When to Ask for Information

* At the start of the conversation
* After answering initial questions
* When visitor asks for a callback
* Before providing detailed information

### What Happens to Leads

Captured leads are saved and can:

* Appear in your conversations list
* Trigger workflows
* Sync to your CRM via integrations
* Generate notifications

## Viewing Agent Conversations

See all interactions your agent has with visitors.

### Accessing Conversations

1. Go to **Agents**
2. Click the **Conversations** tab
3. Browse all visitor chats

### What You Can See

* Full conversation history
* Visitor information (if collected)
* When conversations happened
* How the agent responded

### Using Conversations to Improve

* Find questions the agent couldn't answer
* Add missing information to your knowledge base
* Update your agent's instructions
* Identify common visitor needs

## System Prompt Best Practices

Your system prompt tells the agent how to behave.

### What to Include

**Role Definition:**

```
You are a helpful customer support assistant for [Company Name].
Your job is to answer questions about our products and services.
```

**Tone and Style:**

```
Be friendly and professional. Use simple language.
If you don't know something, say so honestly.
```

**Boundaries:**

```
Only answer questions about our company and products.
Don't discuss competitors or unrelated topics.
Never share confidential information.
```

**Fallback Behavior:**

```
If you can't answer a question, offer to connect the
visitor with a human representative.
```

### Example Support Agent Prompt

```
### Role
You are a customer support agent for [Company Name].
Your job is to help visitors with questions about
our products and services.

### Guidelines
- Be friendly, helpful, and professional
- Answer questions based on the knowledge provided
- Keep responses concise and clear
- If you don't know something, say so

### Boundaries
- Only discuss our company and products
- Don't make promises about pricing or availability
- Offer to connect with humans for complex issues

### Collecting Information
When appropriate, ask for the visitor's name and
email so we can follow up.
```

## Testing Your Agent

### Preview Mode

1. Go to your agent settings
2. Click **Preview** or **Test**
3. Chat with your agent as a visitor would

### What to Test

* Does the welcome message appear correctly?
* Are responses accurate and helpful?
* Does it handle unknown questions gracefully?
* Is lead capture working?
* Does the design look right?

### Testing on Your Website

After installing:

1. Visit your website in a new browser/incognito window
2. Click the chat button
3. Have a conversation
4. Check that everything works as expected

## Managing Multiple Agents

Create different agents for different purposes.

### Use Cases for Multiple Agents

* **Support Agent:** Answer product questions
* **Sales Agent:** Qualify leads and book demos
* **FAQ Agent:** Handle common questions
* **Different Languages:** Separate agents for each language
* **Different Products:** Specialized agents per product line

### Managing Your Agents

1. Go to **Agents**
2. See all your agents in the list
3. Each shows:
   * Name and avatar
   * Linked employee
   * Status (active/inactive)
   * Conversation count

### Duplicating an Agent

To create a similar agent:

1. Find the agent to copy
2. Click the **three-dot menu** (⋮)
3. Select **Duplicate**
4. Edit the copy as needed

## Troubleshooting

### "Agent not appearing on website"

* Verify the embed code is installed correctly
* Check for JavaScript errors in browser console (F12)
* Make sure the code is before `</body>`
* Clear your website's cache
* Check if ad blockers are interfering

### "Agent looks wrong"

* Check appearance settings in your agent config
* Look for CSS conflicts with your website
* Try adjusting z-index if elements overlap
* Verify the embed code is current

### "Responses are incorrect"

* Review and update your knowledge base
* Adjust the AI employee's instructions
* Add more relevant documents
* Refine the system prompt

### "Lead capture not working"

* Check lead capture settings are enabled
* Verify required fields are configured
* Test in preview mode first
* Check for form validation errors

### "Agent is slow to respond"

* This is normal - AI takes a moment to generate responses
* Longer responses take more time
* Complex knowledge base searches may be slower
* Response times vary with server load

## Best Practices

### 1. Write a Clear Welcome Message

* Be friendly and brief
* Tell visitors what the agent can help with
* Set expectations

**Good:** "Hi! I can help with product questions, pricing, and support. What can I assist with?"

**Too Long:** "Welcome to our website! This is an AI-powered assistant that can help you with a wide variety of questions about our products, services, pricing, shipping, returns, and more. Please feel free to type your question below."

### 2. Keep Knowledge Base Updated

Your agent is only as good as its knowledge:

* Add new information regularly
* Remove outdated content
* Include FAQs and common questions
* Update when products or pricing change

### 3. Monitor and Improve

Review conversations weekly:

* Note questions the agent couldn't answer
* Find patterns in what visitors ask
* Update based on real visitor needs
* Adjust the system prompt as needed

### 4. Set Realistic Expectations

* The agent can't do everything
* Have a way for visitors to reach humans
* Be clear about limitations in your welcome message

### 5. Test Regularly

* Test after any changes
* Try different types of questions
* Check on different devices
* Verify embed code still works after website updates

## Frequently Asked Questions

**Q: Can visitors talk to a human instead?**
A: Configure your agent to offer human handoff. You can connect to your [Inbox](/engage/inbox) or other support systems.

**Q: How many agents can I create?**
A: Check your subscription for limits. Most plans allow multiple agents.

**Q: Does it work on mobile?**
A: Yes! Agents are fully responsive and work on all devices.

**Q: Can I see who my visitors are?**
A: You can see information visitors provide through lead capture. Anonymous visitors show limited details.

**Q: Is there a limit to conversations?**
A: Conversations use your subscription's message allowance.

**Q: Can I use different AI employees for different agents?**
A: Yes! Each agent can be powered by a different [AI Employee](/overview/employees).

**Q: Can I have the agent only appear on certain pages?**
A: Yes, only add the embed code to pages where you want the agent.

## Next Steps

* [Create AI Employees](/overview/employees)
* [Build Your Knowledge Base](/overview/knowledge-base)
* [Set Up Integrations](/settings/integrations)
* [Manage Your Inbox](/engage/inbox)


# WhatsApp
Source: https://docs.parallellabs.app/inbound/agents/whatsapp

Connect an agent to WhatsApp so customers can message your business directly.

Let customers message your agent on WhatsApp and get answers instantly. WhatsApp runs on Twilio — the same connection that powers your SMS and voice channels.

## How It Works

When someone messages your WhatsApp business number:

1. Twilio receives the message and forwards it to your agent
2. Your agent reads it, checks your knowledge base, and writes a reply
3. The reply goes back to the customer on WhatsApp

Conversations appear alongside every other channel, so you can read the full history and step in yourself whenever you want.

## Before You Start

You need two things:

* **A Twilio integration** connected under Settings → Integrations. If you already send SMS or make calls, you're done — WhatsApp reuses that same connection.
* **A WhatsApp sender registered on your Twilio account.** WhatsApp numbers are approved by Meta, so this part happens in the Twilio Console. The next section walks through it.

> **One number, one purpose:** A phone number registered with WhatsApp can't already be in use with another WhatsApp account, and each number can power only one agent at a time.

## Step 1: Register a WhatsApp Sender in Twilio

This happens once per phone number, in the Twilio Console.

### What You'll Need

* An upgraded (paid) Twilio account
* Administrator access to your company's Meta Business portfolio — or you can create one during signup
* A phone number that isn't already registered with WhatsApp, and isn't attached to an IVR/phone menu. It must be able to receive an SMS or a voice call for verification.

### The Steps

1. In the Twilio Console, go to **Messaging → Senders → WhatsApp Senders**
2. Click **Create new sender**
3. Pick the phone number you want to use — either a Twilio number or your own — and click **Continue**
4. Click **Continue with Facebook**. A Meta signup window opens.
5. Sign in to Facebook and follow the prompts to connect your WhatsApp Business Account
6. Verify the phone number by SMS or voice call, and enter the code

Complete every step in the same browser window. Meta and Twilio exchange information as you go, and switching browsers partway through breaks the handoff.

### Choosing a Display Name

Your display name is what customers see above your messages. Meta reviews it against their display name guidelines. If your name is rejected, the number is capped at 250 business-initiated messages per 24 hours until it's fixed — so it's worth getting right the first time.

Approval is usually quick, but if Meta also needs to verify your business, that review can take several weeks depending on your region.

## Step 2: Connect the Sender to Your Agent

Once Twilio shows your sender as registered:

1. Go to **Agents** and open the agent you want to answer WhatsApp messages
2. Open the **WhatsApp** tab
3. Turn on **Enable WhatsApp Agent**
4. Pick your **Twilio Integration** from the dropdown
5. Pick your **WhatsApp Sender** from the list of numbers on that Twilio account
6. Save the agent

Selecting a sender automatically points its incoming-message webhook at your agent — there's nothing to copy or paste into Twilio. You'll see a brief "Configuring webhook" message while that happens.

If the sender list is empty, the Twilio account doesn't have an approved WhatsApp sender yet. Go back to Step 1. Use the refresh button beside the dropdown after registering one.

## Step 3: Test It

Message your WhatsApp business number from your own phone. Your agent should reply within a few seconds.

If nothing comes back, see Troubleshooting below.

## Settings

| Setting                             | What It Does                                                                                            |
| ----------------------------------- | ------------------------------------------------------------------------------------------------------- |
| **Enable WhatsApp Agent**           | Turns the channel on or off                                                                             |
| **Twilio Integration**              | Which Twilio account the number belongs to                                                              |
| **WhatsApp Sender**                 | The number customers message                                                                            |
| **Auto-reply to incoming messages** | When off, messages are still recorded but the agent won't answer — useful if you'd rather reply by hand |
| **Custom Channel Instructions**     | Advanced. Replaces the default WhatsApp instructions, e.g. to change tone or response length            |

### Formatting

WhatsApp doesn't render Markdown, so your agent formats replies using WhatsApp's own conventions — `*bold*` and `_italic_` — and writes links as plain URLs. Messages are capped at 4,096 characters.

## The 24-Hour Reply Window

WhatsApp limits when businesses can send free-form messages. **After a customer messages you, you have 24 hours to reply freely.** The clock resets every time they send something new.

For your agent, this is rarely a problem: it answers immediately, well inside the window.

It matters for **manual replies**. If you open an old WhatsApp conversation and try to reply more than 24 hours after the customer's last message, the send will fail. This is a WhatsApp rule, not a platform limit. Sending outside that window requires a message template pre-approved by Meta, which this channel doesn't currently support.

**In practice:** reply while the conversation is fresh. If the window has closed, wait for the customer to message again — that reopens it.

## Testing Without an Approved Sender

If you want to try WhatsApp before finishing Meta approval, Twilio offers a sandbox with a shared test number.

1. Open the Try WhatsApp page in the Twilio Console and confirm the terms
2. Send `join <your-code>` from your phone to the sandbox number to connect it
3. Under **Sandbox settings → Sandbox configuration**, set **"When a message comes in"** to your webhook URL, ending in `/webhooks/twilio/whatsapp`

The sandbox is configured separately from a registered sender, so the agent's sender dropdown won't manage it — you set that webhook yourself. Sandbox sessions expire three days after joining, and only phones that have joined your sandbox can receive messages. It's for testing only; move to a registered sender for real customers.

## Viewing Conversations

WhatsApp conversations appear under **Conversations**, tagged with the WhatsApp channel. You can:

* Read the full back-and-forth
* See the customer's phone number and WhatsApp profile name
* Send a manual reply (subject to the 24-hour window above)

The agent saves the sender's name and number automatically, so repeat messages from the same person stay in one thread.

## Troubleshooting

**The sender dropdown is empty**
The Twilio account has no approved WhatsApp sender. Finish Step 1, then click refresh beside the dropdown.

**Messages arrive but the agent never replies**
Check that **Enable WhatsApp Agent** and **Auto-reply** are both on. If auto-reply is off by design, replies must be sent by hand from Conversations.

**Nothing arrives at all**
Re-select the sender in the WhatsApp tab to reconfigure its webhook. In the Twilio Console, the sender's incoming-message webhook should point at your site followed by `/webhooks/twilio/whatsapp`.

**A number is greyed out in the dropdown**
Another agent is already using it. Each WhatsApp number can serve only one agent — free it up on the other agent first.

**A manual reply failed to send**
The 24-hour window has probably closed. Wait for the customer to message again.

**Replies are cut off**
WhatsApp caps messages at 4,096 characters and longer replies are trimmed. Use Custom Channel Instructions to ask for shorter answers.

## Related Pages

* [Agents](/inbound/agents) — building and configuring agents
* [Integrations](/settings/integrations) — connecting Twilio and other services
* [Inbox](/engage/inbox) — managing conversations across channels


# App Builder
Source: https://docs.parallellabs.app/inbound/app-builder

Create custom web apps with AI, connect them to your data, and publish them.

Build custom web apps by describing what you want. The AI writes the code, connects it to your data, and shows a live preview — then you publish it, optionally to your own domain.

## What Is App Builder?

App Builder lets you create full web apps — not just single pages — through an AI coding assistant. Think internal CRMs, support inboxes, client portals, booking pages, job boards, or onboarding checklists. You describe features in chat; the AI builds them. The result is standard, exportable code you can download and run anywhere.

## Creating an App

1. Select your company from the sidebar
2. Click **App Builder** (under Inbound)
3. Start from a prompt — describe the app you want, for example:
   * "An internal CRM with companies, contacts, and deal stages"
   * "A support inbox that tracks tickets to resolution"
   * "A booking page where customers schedule calls"
4. Or click **New App** and set a **Name**, **Description**, **Icon**, and **App Type**:
   * **Internal app** — runs inside the platform for your team, using your existing login
   * **External app** — a standalone, public app anyone can sign up for, with full theming
5. The builder opens so you can refine it

## Building in the Editor

The editor has three panes:

* **Chat (left)** — ask for what you need: "Add a table of leads with search," "Create a form to add a contact," "Show a chart of deals by stage." The AI writes the code and the preview updates.
* **Files (center)** — the project's files. Click any file to view or edit the code directly.
* **Preview (right)** — a live, running view of your app that refreshes as changes are made.

You can keep multiple chat threads per app to work on different features. Click **Save** to keep your changes.

## Connecting to Your Data

Apps can read and write your platform **tables** (including Smart Lists) so they work with your real data:

* Auto-generated data helpers are included in each app, so the AI can wire up reads and writes for you.
* Just ask in chat — for example, "Load rows from my Leads table into this view," or "Save this form to my Contacts table."

Every published app also exposes a connection point so external AI tools can work with its data securely.

## Publishing

1. In the editor, click **Publish**
2. The app builds and deploys; the status moves to **Published / Live**
3. Internal apps become available to your company; external apps get a public URL

Publish again any time to push a new version.

### Custom Domains

External apps can run on your own domain:

1. Open the app's **Custom Domain** option
2. Enter your domain and follow the DNS instructions
3. Once the domain shows **Active**, your app is live there

## Sharing, Export, and More

* **Sharing** — keep an app **Private**, **Shared** with your company, or **Public** via link.
* **Export** — download the full source code as a standard project. You can run `npm install` and `npm run build` and host it anywhere; nothing ties it to the platform.
* **Git** — link an app to a GitHub repository to push and pull changes.
* **Themes** (external apps) — customize colors, fonts, and styling to match your brand.
* **Monetization** (external apps) — give each user starting credits, charge per action, and sell credit packs.

## App Builder vs. Apps

* **App Builder** is where you *create and edit* apps.
* **[Apps](/apps)** is where you *open and use* the apps you've published.

## Next Steps

* [Open your published apps](/apps)
* [Build a Landing Page](/inbound/landing-pages)
* [Work with your data in Smart Lists](/find-enrich/smart-lists)


# Landing Pages
Source: https://docs.parallellabs.app/inbound/landing-pages

Build and publish marketing and lead-capture pages with an AI-assisted builder.

Build marketing and lead-capture pages with an AI-assisted builder — describe what you want, watch it appear, and publish to your own domain.

## What Are Landing Pages?

Landing pages are single web pages for campaigns, sign-ups, and lead capture. You build them by chatting with an AI builder rather than writing code, connect a form to a Smart List to collect leads, and publish to a platform URL or your own custom domain.

## Creating a Landing Page

1. Select your company from the sidebar
2. Click **Landing Pages** (under Inbound)
3. Click **Create New Landing Page**
4. Fill in the details:
   * **Page Name** — e.g., "Summer Campaign"
   * **Description** — optional
   * **Brand Kit** — optional; applies your saved logo, colors, and styles
   * **Lead List** — optional; choose (or create) the Smart List that captures form submissions
5. Click **Create** to open the builder

## Building and Editing

The builder has three panels:

* **Chat (left)** — describe changes in plain language: "Add a headline and a sign-up form," "Make the buttons blue," "Add a testimonials section." The AI updates the page as you go.
* **Preview (center)** — a live view of the page that updates with each change.
* **Settings (right)** — edit the name, description, brand kit, lead list, and publish domain.

Click **Save** to keep your changes.

## Capturing Leads

If you attach a **Lead List** to a page, form submissions are saved as rows in that Smart List — ready for [Sequences](/engage/sequences), [Marketing Campaigns](/engage/marketing-campaigns), or any list action. Add or change the lead list any time from the builder's settings panel.

## Publishing

1. In the builder, click **Publish**
2. The page goes live and its status changes to **Published**
3. Share the public URL, or connect a custom domain (below)

To take a page down, use **Unpublish** — it returns to draft without deleting your work. Republish whenever you're ready.

### Custom Domains

You can publish to your own domain instead of the default URL:

1. Open the **Domains** tab and click **Add Domain**
2. Follow the DNS instructions and wait for the domain to show **Verified**
3. In a page's settings, choose that domain as the **Publish Domain**
4. Publish — your page is now live on your own domain

## Tips

* **Start with the goal.** Tell the builder what the page is *for* ("a webinar sign-up with name and email") before fine-tuning the design.
* **Attach a lead list early** so you never lose a submission while testing.
* **Preview on mobile.** Check how the page looks at a narrow width before publishing.

## Next Steps

* [Build an App](/inbound/app-builder)
* [Capture leads into a Smart List](/find-enrich/smart-lists)
* [Follow up with a Sequence](/engage/sequences)


# Brand Kit
Source: https://docs.parallellabs.app/inbound/landing-pages/brand-kit

Company-wide assets — favicon, logo, images, CSS — that are injected into your published landing pages automatically.

A brand kit holds the pieces every page should share. Set them once, and pages you build pick them up without you re-uploading a logo or re-pasting an analytics snippet each time.

You can keep several kits, so different campaigns or different clients can each have their own branding. One kit is marked **Default**.

## Creating a Kit

1. Select your company from the sidebar
2. Go to **Landing Pages** → **Brand Kit**
3. Click **Create Brand Kit**
4. Give it a **Name** — for example, "Main Brand" or "Campaign 2024"
5. Save, then fill in its contents

## What Goes in a Kit

| Section                       | What it's for                                                          |
| ----------------------------- | ---------------------------------------------------------------------- |
| **Favicon**                   | The small icon in the browser tab                                      |
| **Logo**                      | Your logo, available to every page                                     |
| **Custom Images**             | Any other reusable imagery                                             |
| **Brand Guidelines**          | Plain-language rules the AI builder follows                            |
| **Global CSS** *(Advanced)*   | Styling applied across pages                                           |
| **Head Scripts** *(Advanced)* | Code injected into every page's `<head>` — analytics tags, for example |

### Image descriptions

Each custom image can carry a description written for the AI — something like "Works well with dark backgrounds." When you ask the builder for a page, it uses these descriptions to choose sensibly between your images instead of picking at random.

Descriptive notes here save a lot of correcting later.

### Brand guidelines

This is where you tell the builder how pages should look and feel — colors, spacing, tone, what to avoid. The more concrete the guidance, the less you'll be correcting the same thing on every page.

### Head scripts

Anything that belongs in every page's `<head>` goes here: analytics, tag managers, verification tags. Adding it once to the kit applies it everywhere, so you can't forget it on a new page.

## Using a Kit

Kits apply to [landing pages](/inbound/landing-pages) when they're published. Different pages can use different kits, which is what makes multiple kits worthwhile — one per brand or client, rather than one per page.

## Tips

1. **Fill in the descriptions.** They're the difference between the AI picking the right image and picking any image.
2. **Put analytics in Head Scripts**, not in individual pages — one place to update, nothing missed.
3. **Write real brand guidelines.** "Use plenty of whitespace, avoid stock-photo people, headline font is bold" changes output; "look professional" doesn't.
4. **One kit per brand**, not per campaign, unless the campaigns genuinely look different.

## Next Steps

* [Build a landing page](/inbound/landing-pages)
* [Set up a custom domain](/inbound/landing-pages/tracking-domains)
* [Brand Images for generated visuals](/engage/content-engine/brand-images)


# Domains
Source: https://docs.parallellabs.app/inbound/landing-pages/tracking-domains

Publish your landing pages to your own branded URL.

Landing pages are served from a platform URL unless you attach a custom domain. Using your own — something like `pages.yourbrand.com` — keeps the link consistent with your brand, which matters when the URL appears in an ad or an email.

## Adding a Domain

1. Select your company from the sidebar
2. Go to **Landing Pages** → **Domains**
3. Click **Add Domain**
4. Enter the domain or subdomain you want to use
5. Copy the **A Record (IP Address)** from the table

Then, at your DNS provider:

1. Open your domain's DNS settings
2. Create an **A record**
3. Point it at the IP address shown
4. Save and wait for DNS to propagate — up to 24–48 hours
5. Return here and click the refresh icon to **Verify DNS & SSL**

SSL is provisioned automatically once DNS resolves correctly.

## Domain Status

| Status         | Meaning                     |
| -------------- | --------------------------- |
| **Pending**    | Waiting for DNS             |
| **Verified**   | DNS is correct              |
| **SSL Active** | Live and secured over HTTPS |
| **Failed**     | Hover the status to see why |

## Using a Subdomain

A subdomain such as `pages.yourbrand.com` or `go.yourbrand.com` is usually the right choice. It leaves your main site untouched, and you can add the record without touching the DNS your primary website depends on.

## Limits

There's a per-company limit on custom domains. At the cap, the add button is disabled and the limit is shown.

## Troubleshooting

**Stuck on Pending.** Check the A record points at exactly the IP in the table, and that it's an A record rather than a CNAME.

**Failed.** Hover for the error. Usually a wrong IP or a leftover conflicting record for the same hostname.

**Certificate warnings.** These resolve once the status reaches **SSL Active**. If DNS only just propagated, give provisioning a few minutes.

## Next Steps

* [Build a landing page](/inbound/landing-pages)
* [Set up a Brand Kit](/inbound/landing-pages/brand-kit)
* [Domains for full websites](/inbound/websites/website-domains)


# Sites, Pages, or Apps?
Source: https://docs.parallellabs.app/inbound/overview

Everything customers come to: chatbots that answer them, sites and pages that convert them, and apps you build for them.

The rest of the platform reaches out. This section is what's waiting when someone arrives — an agent that answers at 2am, a landing page behind an ad, a full website, or a custom app.

## In This Section

* **[Agents](/inbound/agents)** — Chatbots on your website, by phone, and in messaging apps that answer questions and capture leads.
* **[WhatsApp](/inbound/agents/whatsapp)** — Connect an agent to WhatsApp so customers can message your business directly.
* **[Websites](/inbound/websites)** — Build complete marketing websites with a built-in content management system.
* **[Landing Pages](/inbound/landing-pages)** — Single pages for campaigns and lead capture, built with an AI chat builder.
* **[App Builder](/inbound/app-builder)** — Create custom web apps with AI, connect them to your data, and publish them.

## Websites, Landing Pages, or Apps?

All three are built by describing what you want to an AI builder, and all three publish to a custom domain. They differ in what they're for:

|                                             | Best for                 | Shape                                  |
| ------------------------------------------- | ------------------------ | -------------------------------------- |
| **[Websites](/inbound/websites)**           | Your main marketing site | Many pages, navigation, a built-in CMS |
| **[Landing Pages](/inbound/landing-pages)** | One campaign, one offer  | A single focused page                  |
| **[App Builder](/inbound/app-builder)**     | Something interactive    | An app with data behind it             |

If it needs a menu and more than a few pages, build a website. If it's the destination for one ad or one email, build a landing page. If people need to *do* something rather than read something, build an app.


# Websites
Source: https://docs.parallellabs.app/inbound/websites

Build a complete marketing website — pages, copy, images, and a built-in content management system — by describing it.

Describe the site you want and it gets built: pages, navigation, copy, and images, with a content management system behind it for anything that repeats, like blog posts or case studies. No coding necessary.

> **Beta.** Websites is a newer feature and is still being refined.

## Creating a Website

1. Select your company from the sidebar
2. Go to **Websites**
3. Click **New website** (or use the prompt box on the first-run screen)
4. Describe the site you want — for example, "A site for my coffee shop with a menu and events"
5. Decide whether to leave **Use company context** on
6. Click **Create**

The AI names the site, builds the pages, and wires up the navigation.

**Use company context** gives the builder your company name, website, and description so it writes real copy about your actual business. Turn it **off** when you're building a site for a client, so it doesn't write about you instead of them.

## The Editor

Opening a site puts you in a full-screen editor: a chat panel on the left and your site on the right.

Describe changes in the chat and watch them appear. The top bar switches the right-hand panel between:

* **Preview** — the site as visitors see it
* **Code** — the underlying files
* **Dashboard** — settings, content, and domains

You'll also find **File history**, **Refresh preview**, and a **Company context** switch in the top bar.

## The Content Management System

The CMS is what separates a website from a landing page. It's for content that repeats — blog posts, case studies, team members, products.

### Collections and entries

A **collection** is a type of repeating content. Each item in it is an **entry**.

Open **Dashboard** → **Content** to manage them. If you have no collections yet, you can create one, or use **Ask AI to build a blog** to have the builder create a blog with its templates and index page for you.

### Creating a collection

A collection has:

* **Name** — what the content is, e.g. `posts`
* **Route prefix** — where entries are served, so `/blog` gives you `/blog/<slug>`
* **Item template** — the project file each entry renders through
* **Description**

If the item template doesn't exist yet, ask the AI in the chat to create it.

Templates refer to a collection by its **Name**, not its route prefix — a collection named `posts` is looped over as `{% for post in collections.posts %}`. Using the route prefix instead is the usual reason a listing page renders empty. The collection dialog shows the exact syntax for the name you've typed.

### Creating an entry

Each entry has a **Title**, **Slug**, **Body**, **Excerpt**, **Meta description**, **Hero image URL**, **Author**, **Tags**, and **Status**.

The body can be written in Markdown — toggle the Markdown option and it's converted to HTML when you save.

Entries are listed with their **Title**, **Status**, and **Route**, so you can see exactly where each one is published.

## Publishing

From a site's menu, choose **Publish**. This builds the site and hosts it publicly. Later changes need **Publish a new version**, which rebuilds every page and updates the live site — editing alone doesn't change what visitors see.

Sites show a **Published** or **Draft** chip, and a published site has a live URL you can open. If a publish fails, the menu shows the error with a **Copy error** button.

## Custom Domains

Published sites can be served from your own domain. Use **Add custom domain** from a site's menu, or manage them all under **Websites** → **Domains**.

You must publish a site before attaching a domain to it. See [Domains](/inbound/websites/website-domains) for the DNS setup.

## Tips

1. **Describe the whole site up front.** "A five-page site for a dental practice with services, team, and booking" gives the builder more to work with than "a website."
2. **Turn off company context for client work**, or the copy will be about your business.
3. **Use a collection for anything repeating.** Ten hand-built blog pages become unmaintainable; a collection with ten entries doesn't.
4. **Remember to republish.** Edits are not live until you publish a new version.
5. **Iterate in small steps** in the chat — one change at a time is easier to judge and undo.

## Next Steps

* [Set up a custom domain](/inbound/websites/website-domains)
* [Build a single landing page instead](/inbound/landing-pages)
* [Create a custom app](/inbound/app-builder)


# Domains
Source: https://docs.parallellabs.app/inbound/websites/website-domains

Serve your published websites from your own branded URL.

By default a published site lives on a platform URL. Adding a custom domain serves it from your own address instead — `www.yourbrand.com` rather than a generated link.

## Adding a Domain

1. Select your company from the sidebar
2. Go to **Websites** → **Domains**
3. Click **Add Domain**
4. Enter the domain you want to use
5. Copy the **A Record (IP Address)** shown in the table

Then, at your domain registrar or DNS provider:

1. Open your domain's DNS settings
2. Create an **A record**
3. Point it at the IP address shown in the table
4. Save, and wait for DNS to propagate — this can take 24–48 hours
5. Back in the platform, click the refresh icon to **Verify DNS & SSL**

SSL certificates are provisioned automatically once DNS is pointing correctly. You don't need to buy or upload one.

## Domain Status

| Status         | Meaning                                               |
| -------------- | ----------------------------------------------------- |
| **Pending**    | Waiting for DNS to point at the IP address            |
| **Verified**   | DNS is correct                                        |
| **SSL Active** | DNS is correct and HTTPS is live — ready to use       |
| **Failed**     | Something went wrong; hover the status for the reason |

**SSL Active** is the state you want. Until then, the site is still reachable at its platform URL.

## Publish First

A domain can only be attached to a site that has been **published**. Publish the site, then attach the domain.

## Limits

Each company can have a set number of custom domains. When you reach it, the add button is disabled and a message tells you the maximum.

## Troubleshooting

**Still Pending after a day.** DNS propagation is genuinely slow sometimes. Confirm the A record points at exactly the IP shown, with no typos, and that you created an **A** record rather than a CNAME.

**Failed status.** Hover it for the specific error. The most common causes are a record pointing at the wrong address, or a conflicting record for the same hostname left over from a previous host.

**Works without www but not with it (or the reverse).** Each hostname needs its own record. `example.com` and `www.example.com` are separate.

## Next Steps

* [Publish a website](/inbound/websites)
* [Set up domains for landing pages](/inbound/landing-pages/tracking-domains)


# Automate Browser Tasks with Parallel AI's AI Agent
Source: https://docs.parallellabs.app/integrations/browser-automation

Submit natural language browser tasks to Parallel AI for async AI-driven execution. Supports standard, authenticated, and residential sessions.

## What is Browser Automation?

Parallel AI can execute browser tasks described in plain natural language. You submit a task description, receive a task ID, and poll for results when they're ready. This lets you automate complex web interactions without writing brittle selectors or scripts.

## Session Types

Choose the session type that fits your use case:

| Type            | Description                                                                             |
| --------------- | --------------------------------------------------------------------------------------- |
| `regular`       | Standard browser session (default). No login state or special routing.                  |
| `authenticated` | Uses a saved integration (LinkedIn, Twitter, etc.) to run tasks in a logged-in session. |
| `residential`   | Routes traffic through a residential proxy. Specify a US ZIP code for geo-targeting.    |

## Submit a Task

Send a `POST` request to `/api/v0/browser-task` with your task description and session type.

```bash theme={null}
curl -X POST "https://api.parallellabs.app/api/v0/browser-task" \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "task": "Go to https://example.com and return the page title",
    "sessionType": "regular"
  }'
```

**Response:**

```json theme={null}
{
  "taskId": "task_abc123",
  "status": "pending",
  "message": "Task submitted successfully"
}
```

<Tip>
  Save the `taskId` from the response — you'll need it to poll for results or cancel the task.
</Tip>

## Poll for Results

Retrieve the current status and result of a task by calling `GET /api/v0/browser-task/{task_id}`.

```bash theme={null}
curl "https://api.parallellabs.app/api/v0/browser-task/task_abc123" \
  -H "X-API-Key: YOUR_API_KEY"
```

**Possible status values:**

| Status      | Meaning                                 |
| ----------- | --------------------------------------- |
| `pending`   | Task is queued and waiting to start     |
| `running`   | Task is actively being executed         |
| `completed` | Task finished successfully              |
| `failed`    | Task encountered an unrecoverable error |
| `cancelled` | Task was cancelled before completion    |

**Example completed response:**

```json theme={null}
{
  "taskId": "task_abc123",
  "status": "completed",
  "result": {
    "success": true,
    "result": "The page title is 'Example Domain'",
    "creditsCharged": 5,
    "history": [
      { "action": "navigate", "url": "https://example.com", "timestamp": "2024-01-15T10:23:01Z" },
      { "action": "extract", "field": "title", "value": "Example Domain", "timestamp": "2024-01-15T10:23:03Z" }
    ]
  }
}
```

<Note>
  Poll at a reasonable interval (e.g. every 2–5 seconds) to avoid unnecessary API calls. Most tasks complete within 30 seconds.
</Note>

## Cancel a Task

Send a `DELETE` request to cancel a task. Cancellation is only possible while the task is in `pending` or `running` status.

```bash theme={null}
curl -X DELETE "https://api.parallellabs.app/api/v0/browser-task/task_abc123" \
  -H "X-API-Key: YOUR_API_KEY"
```

<Warning>
  Tasks that have already reached `completed`, `failed`, or `cancelled` status cannot be cancelled again.
</Warning>

## Authenticated Sessions

Use `authenticated` sessions to run tasks inside accounts you've connected via Parallel AI integrations (e.g. LinkedIn, Twitter).

<Steps>
  <Step title="List available integrations">
    First, retrieve your saved integrations to find the correct `integrationId`:

    ```bash theme={null}
    curl "https://api.parallellabs.app/api/v0/browser-integrations" \
      -H "X-API-Key: YOUR_API_KEY"
    ```
  </Step>

  <Step title="Submit with an integrationId">
    Pass `sessionType: "authenticated"` and the `integrationId` in your task request:

    ```json theme={null}
    {
      "task": "Check my LinkedIn notifications",
      "sessionType": "authenticated",
      "integrationId": "integration_id_here"
    }
    ```
  </Step>
</Steps>

## Residential Proxy

Route your browser session through a residential IP address. Provide a US ZIP code to target a specific location:

```json theme={null}
{
  "task": "Search Amazon for laptops and return the first 5 results",
  "sessionType": "residential",
  "zipcode": "10001"
}
```

<Tip>
  Residential sessions are useful for tasks that may be geo-restricted or that require a non-datacenter IP to avoid bot detection.
</Tip>

## Vision Mode

Enable `useVision: true` to switch to screenshot-driven automation. In vision mode, the browser agent perceives pages as images rather than DOM elements — useful for visually complex or canvas-based interfaces.

```json theme={null}
{
  "task": "Find and click the blue 'Sign Up' button on the homepage",
  "sessionType": "regular",
  "useVision": true
}
```

<Note>
  For the full Browser Automation API reference, see [Submit a Browser Task](/api-reference/introduction).
</Note>


# Use Claude Code with Parallel AI
Source: https://docs.parallellabs.app/integrations/claude-code

Parallel AI exposes an Anthropic Messages API-compatible endpoint, so Claude Code and any Anthropic SDK can use Parallel AI as their LLM provider.

## Overview

Parallel AI's `/api/v0/claude` endpoints are compatible with the Anthropic Messages API. Point **Claude Code** — or any Anthropic SDK — at Parallel AI and it works out of the box, billed against your Parallel AI credits like any other API usage.

| Endpoint                                       | Description                                                        |
| ---------------------------------------------- | ------------------------------------------------------------------ |
| `POST /api/v0/claude/v1/messages`              | Create a message (streaming and non-streaming, tool use supported) |
| `POST /api/v0/claude/v1/messages/count_tokens` | Estimate token count for a request                                 |
| `GET /api/v0/claude/v1/models`                 | List available models                                              |

<Tip>
  The `model` field accepts **any** Parallel AI model — not just Anthropic
  models — so you can run Claude Code on any model the platform offers.
</Tip>

## Authentication

Authentication works exactly like the rest of the API: send your API key as `Authorization: Bearer <key>` or `x-api-key: <key>`.

Personal Access Keys (`pak_...`) also work but require the `X-Company-ID` header, which Claude Code cannot send by default — use a **Company API Key** for Claude Code.

## Setting up Claude Code

<Steps>
  <Step title="Install Claude Code">
    Skip this if it's already installed.

    ```bash theme={null}
    # macOS / Linux / WSL — native installer
    curl -fsSL https://claude.ai/install.sh | bash

    # Or via npm (requires Node.js 18+)
    npm install -g @anthropic-ai/claude-code
    ```

    On Windows (PowerShell):

    ```powershell theme={null}
    irm https://claude.ai/install.ps1 | iex
    ```

    Verify the install with `claude --version`. See the [official docs](https://code.claude.com/docs) for more install options.
  </Step>

  <Step title="Get a Company API Key">
    In the dashboard, go to **Integrations** → **API Keys** → **Generate New Key**.
  </Step>

  <Step title="Configure Claude Code">
    Either export environment variables in your shell:

    ```bash theme={null}
    export ANTHROPIC_BASE_URL="https://api.parallellabs.app/api/v0/claude"
    export ANTHROPIC_AUTH_TOKEN="YOUR_COMPANY_API_KEY"
    claude
    ```

    Or persist the settings in `~/.claude/settings.json` (applies to all projects) or `.claude/settings.local.json` in a project. Set the model variables to any Parallel AI model names — for example:

    ```json theme={null}
    {
      "env": {
        "ANTHROPIC_BASE_URL": "https://api.parallellabs.app/api/v0/claude",
        "ANTHROPIC_AUTH_TOKEN": "YOUR_COMPANY_API_KEY",
        "ANTHROPIC_MODEL": "qwen3.8-max",
        "ANTHROPIC_DEFAULT_OPUS_MODEL": "kimi-k3",
        "ANTHROPIC_DEFAULT_SONNET_MODEL": "qwen3.8-max",
        "ANTHROPIC_DEFAULT_HAIKU_MODEL": "deepseek-v4-flash-0731"
      }
    }
    ```
  </Step>

  <Step title="Verify">
    Run `/status` inside Claude Code — it should show your custom base URL — or test the endpoint directly:

    ```bash theme={null}
    curl -X POST "https://api.parallellabs.app/api/v0/claude/v1/messages" \
      -H "Authorization: Bearer YOUR_COMPANY_API_KEY" \
      -H "anthropic-version: 2023-06-01" \
      -H "Content-Type: application/json" \
      -d '{"model": "qwen3.8-max", "max_tokens": 64, "messages": [{"role": "user", "content": "Hello"}]}'
    ```
  </Step>
</Steps>

## Model names

The `model` field accepts any model name from `GET /api/v0/models` — any provider, not just Anthropic. When Claude Code sends its default Anthropic model IDs (e.g. `claude-sonnet-4-5-20250929`), they are routed by tier to Parallel AI models:

* **Opus-class** requests run on the platform's highest-capability model
* **Sonnet-class** requests run on the balanced default model
* **Haiku-class** requests (background tasks) run on the fastest low-cost model

Set `ANTHROPIC_MODEL` and the `ANTHROPIC_DEFAULT_*_MODEL` variables to exact Parallel AI model names for full control over routing.

Optionally set `CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERY=1` so Claude Code lists Parallel AI's models in its `/model` picker.

## Using the Anthropic SDK directly

<CodeGroup>
  ```python Python theme={null}
  import anthropic

  client = anthropic.Anthropic(
      base_url="https://api.parallellabs.app/api/v0/claude",
      auth_token="YOUR_COMPANY_API_KEY",
  )

  response = client.messages.create(
      model="kimi-k3",
      max_tokens=1024,
      messages=[{"role": "user", "content": "Hello"}],
  )
  print(response.content[0].text)
  ```

  ```typescript TypeScript theme={null}
  import Anthropic from "@anthropic-ai/sdk";

  const client = new Anthropic({
    baseURL: "https://api.parallellabs.app/api/v0/claude",
    authToken: "YOUR_COMPANY_API_KEY",
  });

  const response = await client.messages.create({
    model: "kimi-k3",
    max_tokens: 1024,
    messages: [{ role: "user", content: "Hello" }],
  });
  console.log(response.content[0].text);
  ```
</CodeGroup>

## Notes and limitations

* **Streaming** is fully supported (Anthropic SSE event format with keep-alive pings).
* **Tool use** is supported: Claude Code's client-side tools (file edits, bash, etc.) are forwarded to the model and executed locally by Claude Code — never on Parallel AI's servers.
* Server-side Anthropic tools (web search, code execution), extended thinking blocks, and prompt caching directives are accepted but ignored.
* Usage is billed per request based on real token counts and appears in your usage dashboard under the chat feature.


# Connect AI Assistants via Parallel AI MCP Integration
Source: https://docs.parallellabs.app/integrations/mcp

Parallel AI's MCP server lets Claude Desktop, Cursor, Windsurf, and other AI tools access your agents, sequences, documents, and more.

## What is MCP?

Model Context Protocol (MCP) is an open standard for connecting AI assistants to external tools and data sources. With Parallel AI's MCP server, your AI assistant can:

* Browse and manage your AI agents
* Create and monitor outreach sequences
* Search your knowledge base documents
* Generate AI content, images, audio, and video
* Execute browser automation tasks
* Access every platform feature via natural language

## MCP Server URL

```
https://api.parallellabs.app/api/v0/mcp
```

<Note>
  This server implements MCP Streamable HTTP Transport (JSON-RPC 2.0). It is
  compatible with `mcp-remote` and other standard MCP clients.
</Note>

## Available Tools

| Tool                          | Description                                                 |
| ----------------------------- | ----------------------------------------------------------- |
| `platform_list_capabilities`  | Browse available API endpoints by category                  |
| `platform_get_action_details` | Get the full spec for an endpoint including required params |
| `platform_execute_action`     | Execute any API endpoint with automatic authentication      |

## Connecting Claude Desktop

<Steps>
  <Step title="Open your Claude Desktop config file">
    Locate the configuration file for your operating system:

    * **macOS:** `~/Library/Application Support/Claude/claude_desktop_config.json`
    * **Windows:** `%APPDATA%\Claude\claude_desktop_config.json`
  </Step>

  <Step title="Add the MCP server">
    Paste the following into your config file, replacing `your_api_key_here` with your actual API key:

    ```json claude_desktop_config.json theme={null}
    {
      "mcpServers": {
        "parallel-ai": {
          "command": "npx",
          "args": ["-y", "mcp-remote", "https://api.parallellabs.app/api/v0/mcp"],
          "env": {
            "API_KEY": "your_api_key_here"
          }
        }
      }
    }
    ```
  </Step>

  <Step title="Restart Claude Desktop">
    Fully quit and relaunch Claude Desktop so it picks up the new configuration.
  </Step>

  <Step title="Test the connection">
    Ask Claude something like:

    * *"List my AI agents"*
    * *"Create a new email sequence for cold outreach"*

    Claude will automatically use the Parallel AI tools to fulfill your request.
  </Step>
</Steps>

## Connecting Cursor / Windsurf

<Steps>
  <Step title="Open the MCP settings">
    Navigate to **Settings → MCP** inside Cursor or Windsurf.
  </Step>

  <Step title="Add a new MCP server">
    Fill in the following details:

    * **URL:** `https://api.parallellabs.app/api/v0/mcp`
    * **Authentication:** Bearer token using your Parallel AI API key
  </Step>

  <Step title="Confirm the tools are available">
    The Parallel AI platform tools will appear in your AI assistant's tool list. You can now reference them directly in your prompts.
  </Step>
</Steps>

## Generic mcp-remote Connection

If you're using any MCP client that supports `mcp-remote`, connect with a single command:

```bash theme={null}
npx mcp-remote https://api.parallellabs.app/api/v0/mcp \
  --header "Authorization: Bearer YOUR_API_KEY"
```

## Authentication for MCP

MCP authentication follows the same pattern as the REST API.

**Company API Key:**

```
Authorization: Bearer YOUR_COMPANY_API_KEY
```

**Personal Access Key:**

```
Authorization: Bearer pak_your_key
X-Company-ID: your_company_id
```

<Tip>
  You can find your API keys on the **Integrations** page (Company API Keys) or
  your **Profile** page (Personal Access Keys) in the Parallel AI dashboard.
</Tip>

## Example MCP Interactions

The three tools work together to let your AI assistant discover and execute any platform action.

<CodeGroup>
  ```json platform_list_capabilities theme={null}
  {
    "tool": "platform_list_capabilities",
    "input": {
      "category": "Sequences"
    }
  }
  ```

  ```json platform_get_action_details theme={null}
  {
    "tool": "platform_get_action_details",
    "input": {
      "operationId": "post_sequences.create_sequence"
    }
  }
  ```

  ```json platform_execute_action theme={null}
  {
    "tool": "platform_execute_action",
    "input": {
      "operationId": "get_agents.list_agents",
      "parameters": {}
    }
  }
  ```
</CodeGroup>

<Note>
  For the full MCP API reference, including all available operation IDs and
  response schemas, see the [MCP Overview](/api-reference/introduction).
</Note>


# Use Parallel AI as an OpenAI API Drop-In Replacement
Source: https://docs.parallellabs.app/integrations/openai-compatibility

Parallel AI's chat completions endpoint is fully OpenAI-compatible. Point any OpenAI SDK at your Parallel AI endpoint and start using it immediately.

## Overview

Parallel AI's `POST /api/v0/chat/completions` endpoint is 100% compatible with OpenAI's Chat Completions API spec. To migrate an existing integration, you only need to change the `base_url` in your client — no other code changes are required.

<Tip>
  This means any library, tool, or service built for the OpenAI API works out of
  the box with Parallel AI. Swap the base URL and your API key, and you're done.
</Tip>

## Quick Start

<CodeGroup>
  ```python Python theme={null}
  from openai import OpenAI

  client = OpenAI(
      api_key="YOUR_API_KEY",
      base_url="https://api.parallellabs.app/api/v0"
  )

  response = client.chat.completions.create(
      model="gpt-4o",
      messages=[
          {"role": "system", "content": "You are a helpful assistant."},
          {"role": "user", "content": "What can Parallel AI help me with?"}
      ]
  )

  print(response.choices[0].message.content)
  ```

  ```javascript Node.js theme={null}
  import OpenAI from 'openai';

  const client = new OpenAI({
    apiKey: 'YOUR_API_KEY',
    baseURL: 'https://api.parallellabs.app/api/v0',
  });

  const response = await client.chat.completions.create({
    model: 'gpt-4o',
    messages: [
      { role: 'system', content: 'You are a helpful assistant.' },
      { role: 'user', content: 'What can Parallel AI help me with?' },
    ],
  });

  console.log(response.choices[0].message.content);
  ```

  ```bash cURL theme={null}
  curl -X POST "https://api.parallellabs.app/api/v0/chat/completions" \
    -H "X-API-Key: YOUR_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{
      "model": "gpt-4o",
      "messages": [
        { "role": "system", "content": "You are a helpful assistant." },
        { "role": "user", "content": "What can Parallel AI help me with?" }
      ]
    }'
  ```
</CodeGroup>

## Supported Parameters

The following request parameters are fully supported:

| Parameter           | Type             | Description                                             |
| ------------------- | ---------------- | ------------------------------------------------------- |
| `model`             | string           | The model to use for completion                         |
| `messages`          | array            | Array of message objects with `role` and `content`      |
| `temperature`       | number           | Sampling temperature between 0 and 2                    |
| `max_tokens`        | integer          | Maximum number of tokens to generate                    |
| `stream`            | boolean          | Stream the response using Server-Sent Events (SSE)      |
| `top_p`             | number           | Nucleus sampling probability mass                       |
| `frequency_penalty` | number           | Penalise repeated tokens based on frequency             |
| `presence_penalty`  | number           | Penalise tokens based on presence in prior text         |
| `tools`             | array            | List of tools the model may call                        |
| `tool_choice`       | string or object | Controls which tool (if any) the model uses             |
| `response_format`   | object           | Set to `{ "type": "json_object" }` to force JSON output |
| `stop`              | string or array  | Up to 4 sequences where the API will stop generating    |
| `seed`              | integer          | Seed for deterministic sampling                         |
| `logit_bias`        | object           | Modify the likelihood of specific tokens appearing      |
| `user`              | string           | A unique identifier for the end user                    |

## Streaming

Set `stream: true` to receive a streaming response using Server-Sent Events (SSE), identical to OpenAI's streaming format.

<CodeGroup>
  ```python Python theme={null}
  from openai import OpenAI

  client = OpenAI(
      api_key="YOUR_API_KEY",
      base_url="https://api.parallellabs.app/api/v0"
  )

  stream = client.chat.completions.create(
      model="gpt-4o",
      messages=[{"role": "user", "content": "Tell me about AI for business."}],
      stream=True,
  )

  for chunk in stream:
      if chunk.choices[0].delta.content is not None:
          print(chunk.choices[0].delta.content, end="", flush=True)
  ```

  ```javascript Node.js theme={null}
  import OpenAI from 'openai';

  const client = new OpenAI({
    apiKey: 'YOUR_API_KEY',
    baseURL: 'https://api.parallellabs.app/api/v0',
  });

  const stream = await client.chat.completions.create({
    model: 'gpt-4o',
    messages: [{ role: 'user', content: 'Tell me about AI for business.' }],
    stream: true,
  });

  for await (const chunk of stream) {
    process.stdout.write(chunk.choices[0]?.delta?.content ?? '');
  }
  ```
</CodeGroup>

<Note>
  Each streamed chunk follows the standard OpenAI SSE format with `data:` prefixed JSON objects, ending with `data: [DONE]`.
</Note>

## Learn More

* [Chat Features Overview](/overview/chat) — explore Parallel AI's chat capabilities
* [Chat Completions API Reference](/api-reference/introduction) — full endpoint documentation


# Parallel AI: Build Agents and Automate Business Growth
Source: https://docs.parallellabs.app/introduction

Parallel AI is an all-in-one AI platform for building agents, automating outreach, generating media, and integrating AI into any business workflow.

## Welcome to Parallel AI

Parallel AI is an all-in-one platform that brings the full power of artificial intelligence to your business — from intelligent agents and automated outreach to media generation and deep workflow integrations. Whether you're a developer building custom automations or a business owner looking to grow faster, Parallel AI gives you the tools to make it happen.

<Note>
  The Parallel AI REST API is available at **`https://api.parallellabs.app`**. All endpoints are served under `/api/v0/`.
</Note>

## Platform Capabilities

### AI Agents

Build and deploy intelligent agents that engage your customers across every channel — all from a single platform.

* **Web** — embed chat widgets directly on your website
* **Email** — automate inbound and outbound email conversations
* **SMS** — reach customers on mobile with AI-driven text messaging
* **Voice** — deploy phone agents that speak and listen naturally

Agents can be connected to your Knowledge Base, given custom instructions, and integrated with external tools to handle complex, multi-step conversations on your behalf.

### OpenAI-Compatible Chat Completions

Parallel AI exposes an OpenAI-compatible `/chat/completions` endpoint, so you can drop Parallel AI into any existing integration that supports the OpenAI SDK — just swap the base URL. Run completions against leading models without changing your code.

### Smart Lists

Manage your contacts with Smart Lists — a powerful contact management system that lets you segment, filter, and target the right people at the right time. Import contacts, apply tags, and build dynamic audiences that automatically update as your data changes.

### Sequences

Automate your outreach with multi-step Sequences across email and SMS. Design drip campaigns, follow-up flows, and nurture journeys that run automatically based on triggers, schedules, or contact behavior — all without manual effort.

### Knowledge Base Documents

Give your agents and workflows access to your own content. Upload documents, FAQs, product guides, and more to the Knowledge Base. Agents use this information to answer questions accurately and consistently, grounded in your source material.

### Media Generation

Generate rich media assets on demand:

* **Text-to-Speech** — powered by ElevenLabs, convert any text into natural-sounding audio with a wide selection of voices
* **Images** — generate custom images from text prompts for marketing, content creation, and more
* **Video** — create AI-generated video content to power your campaigns and communications

### Browser Automation

Automate complex web-based tasks with AI-driven browser sessions. Parallel AI supports:

* **Residential sessions** — browse the web with real residential IP addresses for accurate, unblocked results
* **Authenticated sessions** — interact with logged-in web applications as if you were at your keyboard

Hand off repetitive browser workflows to AI and get back structured results, screenshots, or extracted data.

### MCP Server Integration

Connect Parallel AI to your favorite AI-powered development tools via the **Model Context Protocol (MCP)**. Native integrations are available for:

* **Claude Desktop** — surface Parallel AI capabilities directly inside Claude
* **Cursor** — bring your agents, lists, and sequences into your coding environment
* **Windsurf** — extend your AI-assisted workflows with live Parallel AI data

## Explore the Docs

<CardGroup>
  <Card title="Quickstart" icon="bolt" href="/quickstart">
    Make your first API call in under 5 minutes. Get your API key and run a live request.
  </Card>

  <Card title="Authentication" icon="key" href="/authentication">
    Learn about Company API Keys and Personal Access Keys, and how to use each.
  </Card>

  <Card title="AI Agents" icon="robot" href="/inbound/agents">
    Build multi-channel agents for web, email, SMS, and voice with full customization.
  </Card>

  <Card title="API Reference" icon="code" href="/api-reference/introduction">
    Explore every endpoint in the Parallel AI REST API with live examples.
  </Card>
</CardGroup>


# Chat
Source: https://docs.parallellabs.app/overview/chat

Talk directly with your AI employees to get work done, ask questions, and think out loud.

The Chat feature lets you have conversations with your AI employees. Ask questions, get help with tasks, brainstorm ideas, or just explore what your AI can do.

## Starting a Conversation

1. Select your company from the sidebar
2. Click **Chat**
3. Choose an AI employee to talk to (you'll see their profile cards)
4. Type your message in the text box at the bottom
5. Press **Enter** or click the send button

Your AI employee will respond based on their personality and the knowledge you've given them.

## Choosing the Right AI Employee

Each AI employee has different expertise based on how you've set them up. Choose the one that fits your question:

* **Support-focused employee:** Ask about customer questions, policies, procedures
* **Sales-focused employee:** Get help with pitches, objections, product info
* **Technical employee:** Ask about processes, how things work
* **General assistant:** Good for brainstorming and general questions

**Tip:** You can switch employees mid-conversation if you need a different perspective.

## Sending Messages

### Text Messages

Just type your question or request and press Enter. Be specific for better responses:

| Instead of...       | Try...                                                                   |
| ------------------- | ------------------------------------------------------------------------ |
| "Help me"           | "Help me write an email to a customer who's upset about a late delivery" |
| "What do you know?" | "What do you know about our return policy?"                              |
| "Write something"   | "Write a LinkedIn post about our new product launch"                     |

### Attaching Images

Some AI employees can analyze images:

1. Click the **attachment icon** (paperclip or image icon)
2. Select an image from your computer
3. Add a message explaining what you want to know about the image
4. Send

Example: Attach a screenshot and ask "What's wrong with this design?"

### Uploading Documents

Share documents for context:

1. Click the **upload icon**
2. Select a file (PDF, Word doc, etc.)
3. Ask your question about the document
4. Send

Example: Upload a contract and ask "Summarize the key terms in this agreement."

## Getting Better Responses

### Be Specific

The more context you provide, the better the response:

**Avoid:** "Write an email"
**Better:** "Write a follow-up email to a prospect who attended our webinar last week but hasn't scheduled a demo. Keep it friendly and under 100 words."

### Provide Context

Tell the AI what you're trying to accomplish:

**Avoid:** "What should I say?"
**Better:** "A customer is threatening to cancel because of a billing error. What should I say to keep them as a customer?"

### Ask Follow-Up Questions

The AI remembers the conversation, so build on previous responses:

1. "What are the benefits of our premium plan?"
2. "Now write that as a comparison table"
3. "Make it more concise"

## Conversation Features

### Starting Fresh

To begin a new conversation:

1. Click the **Reset** or **New Chat** button
2. Your conversation history is cleared
3. Start your new conversation

### Viewing Chat History

To see past conversations:

1. Click **Chat History** or the history icon
2. Browse your previous chats
3. Click on one to continue it

### Sharing a Conversation

To share a conversation with a colleague:

1. Click the **Share** button
2. Copy the generated link
3. Send it to whoever needs to see it

They'll be able to view the conversation (but not continue it).

## Adjusting Chat Settings

Click the **Settings** icon (gear) to customize your chat experience:

### Model Selection

Choose which AI model powers your conversation:

* **Fast models:** Quick responses, good for simple questions
* **Advanced models:** Slower but more thoughtful, good for complex tasks

### Context Settings

Control how much context the AI uses:

* More context = better understanding but uses more credits
* Less context = faster and cheaper but may miss nuance

### Connected Tools

Tools let your AI employee take action — not just answer. In the chat settings, click the **Connected Tools** card (it shows how many tools are currently enabled) to turn tools on or off for your conversation.

Built-in tools:

| Tool                          | What It Does                                                                                                                                                                                                    |
| ----------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Enable platform API tools** | Manage lists, sequences, agents, and other platform resources via API                                                                                                                                           |
| **Research Task**             | Delegate complex research tasks to a specialized research agent *(paid plans; +2 credits)*                                                                                                                      |
| **Create Documents**          | Generate PDF, DOCX, or CSV files and email them to you                                                                                                                                                          |
| **Browser Automation**        | Automate a real web browser for web scraping, form filling, and navigating websites — click the tool to choose which browser session it uses                                                                    |
| **Leads & Enrichment**        | Find leads by demographic, firmographic, and buyer-intent filters, and enrich people or companies with contact info, demographics, and social profiles *(availability depends on your plan; uses data credits)* |
| **Delegate Task**             | Hand subtasks to your other AI employees, using their specialized skills and tools                                                                                                                              |

Alongside the built-in tools you'll see your **connected tools** — MCP servers, connected apps, and toolsets set up in [Integrations](/settings/integrations) — each toggled per conversation. From here you can also browse the catalog of 1000+ app integrations (Gmail, Slack, HubSpot, Salesforce, and hundreds more) or click **Manage Toolsets** to group tools into reusable sets.

When the AI uses a tool, the step appears right in the conversation (for example, **Find Leads** or **Enrich Contact**) so you can follow what it's doing.

> Tools require a tool-compatible model — if your current model doesn't support tools, pick a different one in settings. The exact set of tools you see depends on your plan and enabled features.

## Tips for Effective Conversations

1. **One topic at a time** - Don't mix unrelated questions in one message
2. **Use examples** - Show the AI what you want: "Write something like this: \[example]"
3. **Give feedback** - If a response isn't right, explain what to change
4. **Be patient** - Complex questions take longer to answer
5. **Experiment** - Try different ways of asking to find what works best

## Understanding AI Responses

### When the AI Says "I Don't Know"

This usually means:

* The information isn't in the knowledge base
* The question is outside the AI's training
* You need to add more documents to help the AI learn

**Solution:** Add relevant documents to your [Knowledge Base](/overview/knowledge-base).

### When Responses Seem Wrong

AI can make mistakes. If something seems off:

* Ask the AI to double-check its answer
* Verify important information yourself
* Report issues to improve the AI over time

### When Responses Are Too Long/Short

Guide the AI with explicit instructions:

* "Give me a brief 2-sentence answer"
* "Explain this in detail with examples"
* "Summarize in bullet points"

## Frequently Asked Questions

**Q: Does the AI remember previous conversations?**
A: Within a single session, yes. But when you start a new conversation, it begins fresh.

**Q: Who can see my conversations?**
A: By default, only you. Shared conversations can be viewed by anyone with the link.

**Q: How do I make the AI sound more like me/my brand?**
A: Customize the AI employee's personality and instructions in the [Employees](/overview/employees) settings.

**Q: Why is the AI giving wrong information about my company?**
A: The AI only knows what's in your [Knowledge Base](/overview/knowledge-base). Add more documents with accurate information.

**Q: Can I use the AI for customer-facing chat?**
A: Yes! Create an [Agent](/inbound/agents) to embed on your website.

## Next Steps

* [Create AI Employees](/overview/employees)
* [Build Your Knowledge Base](/overview/knowledge-base)
* [Create Customer-Facing Agents](/inbound/agents)


# Dashboard
Source: https://docs.parallellabs.app/overview/dashboard

Your analytics home: activity, usage, and performance across every feature you use.

The Dashboard gives you a bird's-eye view of how your AI employees and campaigns are performing. See key metrics, track usage, and identify what's working.

## Accessing the Dashboard

1. Select a company from the sidebar
2. Click **Dashboard** (it's usually the first item)

## Understanding Your Metrics

### Sequence Performance

If you're running email or LinkedIn sequences, you'll see these key numbers:

| Metric              | What It Means                         |
| ------------------- | ------------------------------------- |
| **Members**         | Total people in your sequences        |
| **Emails Sent**     | Number of emails delivered            |
| **Opens**           | How many people opened your emails    |
| **Clicks**          | How many clicked a link in your email |
| **Replies**         | How many responded                    |
| **Pending Actions** | Actions waiting for your approval     |

**Tip:** Click on any metric card to see more details.

### Open Rate & Click Rate

These percentages help you gauge how engaging your content is:

* **Good open rate:** 20-30%
* **Good click rate:** 2-5%
* **Good reply rate:** 1-3%

If your rates are lower, try:

* Writing more compelling subject lines
* Personalizing your messages
* Testing different send times

## The Usage Chart

The bar chart shows your credit usage over the past 30 days. This helps you:

* See usage trends over time
* Identify busy periods
* Plan for future needs

### Reading the Chart

* **X-axis:** Days of the month
* **Y-axis:** Credits used
* **Bars:** Daily credit consumption

Hover over any bar to see the exact number for that day.

## Feature Usage Breakdown

See which features your team uses most. This pie or bar chart shows usage by feature:

* Chat conversations
* Image generation
* Document processing
* Sequence sends
* And more

This helps you understand where your credits go and what features provide the most value.

## Usage Leaderboard

If you have team members, the leaderboard shows who's using the platform most:

| Column       | What It Shows            |
| ------------ | ------------------------ |
| **User**     | Team member's name       |
| **Usage**    | Their credit consumption |
| **Activity** | Number of actions taken  |

This helps identify power users and ensure the platform is being adopted across your team.

## Pending Approvals

Some sequence actions require your approval before they're sent. You'll see a count of pending items on the Dashboard.

### Reviewing Pending Actions

1. Click **Pending Actions** on the Dashboard
2. Review each message or action
3. Click **Approve** to send it, or **Reject** to cancel

**Why approvals exist:** To ensure AI-generated messages are accurate before they go out to real people.

## Additional Stats

### Smart List Stats

See an overview of your list activity:

* Total lists created
* Total rows across all lists
* Recent list actions

### Inbox Stats

Monitor your email health:

* Emails sent this period
* Emails received
* Bounce rate
* Reply rate

### Content Stats

Track your content creation:

* Content pieces created
* Content published
* Content in draft

## Filtering Your View

### Company vs. Platform View

If you have access to multiple companies, you can switch views:

* **Company:** See metrics for the selected company only
* **Platform:** See combined metrics across all your companies

Use the toggle at the top of the Dashboard to switch.

### Date Ranges

Some charts allow you to change the time period:

* Last 7 days
* Last 30 days
* This month
* Custom range

## Refreshing Data

Dashboard data updates automatically, but if you want the latest numbers:

1. Click the **Refresh** button (usually a circular arrow icon)
2. Wait for the data to reload

## Tips for Using the Dashboard

1. **Check daily** - A quick glance each morning helps you catch issues early
2. **Watch for trends** - Sudden drops in open rates might indicate deliverability issues
3. **Compare periods** - Is this month better than last month?
4. **Act on insights** - If something's not working, make changes

## Frequently Asked Questions

**Q: Why don't I see any data?**
A: If you're new, you may not have any activity yet. Start sending sequences or chatting with AI to generate data.

**Q: How often does the dashboard update?**
A: Most metrics update in real-time or within a few minutes.

**Q: Can I export this data?**
A: Currently, dashboard data is view-only. For detailed reports, check individual features like Sequences for export options.

**Q: Why are my open rates so high (or low)?**
A: Open tracking isn't 100% accurate due to email client privacy features. Focus on trends rather than absolute numbers.

## Next Steps

* [Learn about Sequences](/engage/sequences)
* [Manage Smart Lists](/find-enrich/smart-lists)
* [Check your Inbox](/engage/inbox)


# AI Employees
Source: https://docs.parallellabs.app/overview/employees

Create and customize the AI team members that power chat, agents, and automations.

AI Employees are the heart of the platform. They're your customizable AI assistants that can answer questions, help with tasks, and even interact with your customers. Think of them as team members you can train to know your business inside and out.

## What Is an AI Employee?

An AI employee is a customized AI assistant with:

* A **name and role** (like "Sarah, Customer Support Lead")
* A **personality** (friendly, professional, casual, etc.)
* **Knowledge** (from documents you upload)
* **Special instructions** (how to handle certain situations)

You can have multiple AI employees, each specialized for different tasks.

## Creating Your First AI Employee

1. Select your company from the sidebar
2. Click **Employees**
3. Click **Create Employee** (or the **+** button)
4. Fill in the details:
   * **Name:** What you'll call this AI (e.g., "Alex")
   * **Title:** Their role (e.g., "Sales Assistant")
   * **Photo:** Upload an avatar or use the default
   * **Instructions:** How the AI should behave (more on this below)
5. Click **Save**

That's it! Your AI employee is ready to chat.

## Writing Good Instructions

The instructions (also called "prompt" or "system prompt") tell your AI employee how to behave. This is the most important part of setup.

### What to Include

1. **Role description:** Who is this AI?
2. **Tone and style:** How should they communicate?
3. **Knowledge boundaries:** What should they NOT answer?
4. **Special instructions:** Any specific behaviors

### Example Instructions

**For a Customer Support AI:**

```
You are Alex, a friendly customer support representative for [Company Name].

Your job is to help customers with questions about our products and services. Be warm, patient, and professional.

Important guidelines:
- Always greet customers warmly
- If you don't know an answer, say so and offer to connect them with a human
- Never make up information about products or policies
- Keep responses concise but helpful
- End conversations by asking if there's anything else you can help with
```

**For a Sales Assistant AI:**

```
You are Jordan, a helpful sales assistant for [Company Name].

Your goal is to understand what prospects need and help them see how our products can solve their problems. Be enthusiastic but not pushy.

Guidelines:
- Ask questions to understand their situation
- Focus on benefits, not just features
- Collect their name and email when they show interest
- Never pressure or use aggressive tactics
- If they're not a good fit, be honest about it
```

## Adding a Profile Photo

A friendly face makes your AI more approachable:

1. Click on an existing employee (or create a new one)
2. Click on the avatar/photo area
3. Upload an image from your computer
4. The photo is saved automatically

**Tips:**

* Use a professional-looking photo
* Square images work best
* You can use AI-generated avatars if you prefer

## Editing an AI Employee

To change an existing employee:

1. Go to **Employees**
2. Find the employee you want to edit
3. Click on their card or the **three-dot menu** (⋮)
4. Select **Edit**
5. Make your changes
6. Click **Save**

## Organizing with Tags and Departments

### Adding Tags

Tags help you categorize employees:

1. Edit an employee
2. Find the **Tags** field
3. Type a tag name and press Enter
4. Add as many tags as you need
5. Save

Use tags like: "support", "sales", "technical", "billing", etc.

### Assigning to Departments

If your company has departments set up:

1. Edit an employee
2. Select a **Department** from the dropdown
3. Save

Employees in specific departments may only be accessible to team members in those departments.

## Reordering Employees

Change the order employees appear in the list:

1. Go to **Employees**
2. Hover over an employee card
3. Drag using the handle on the left
4. Drop in the new position

The new order is saved automatically.

## Using the Employee Marketplace

Don't want to start from scratch? Browse pre-made employees:

1. Go to **Employees**
2. Click **Marketplace** or **Browse Templates**
3. Browse available employee templates
4. Click **Install** on one you like
5. It's added to your company

You can then customize the installed employee to fit your needs.

## Deleting an AI Employee

1. Go to **Employees**
2. Click the **three-dot menu** (⋮) on the employee
3. Select **Delete**
4. Confirm the deletion

**Note:** Deleting an employee removes their configuration but doesn't delete conversations you've had with them.

## Tips for Better AI Employees

### 1. Start Simple, Then Refine

Don't try to write perfect instructions on day one. Start with basics, test through conversation, and improve over time.

### 2. Test Thoroughly

After creating an employee, have several conversations to test:

* Does it respond the way you expect?
* Does it handle edge cases well?
* Is the tone appropriate?

### 3. Add Specific Examples

In your instructions, include examples of good responses:

```
When someone asks about pricing, respond like this:
"Great question! Our pricing starts at $29/month for the basic plan. Would you like me to explain what's included?"
```

### 4. Define What NOT to Do

Sometimes it's clearer to say what to avoid:

```
Never:
- Promise features we don't have
- Give refunds without manager approval
- Share customer information
- Use slang or casual language
```

### 5. Update Regularly

As your business changes, update your AI employees' instructions to reflect new products, policies, or procedures.

## Frequently Asked Questions

**Q: How many AI employees can I create?**
A: This depends on your subscription plan. Check your Profile for limits.

**Q: Do AI employees share knowledge?**
A: By default, all employees have access to your Knowledge Base. You can restrict specific documents to specific employees.

**Q: Can I copy an AI employee?**
A: Yes! Click the three-dot menu and select **Duplicate**.

**Q: How do I make an AI employee available on my website?**
A: Create an [Agent](/inbound/agents), which is a customer-facing version of your employee.

**Q: Why does my AI employee give wrong answers?**
A: Check your [Knowledge Base](/overview/knowledge-base). The AI only knows what you've taught it.

## Next Steps

* [Add Documents to Your Knowledge Base](/overview/knowledge-base)
* [Start Chatting](/overview/chat)
* [Create a Website Chatbot](/inbound/agents)


# Knowledge Base
Source: https://docs.parallellabs.app/overview/knowledge-base

Upload the documents, websites, and files that train your AI employees.

Your Knowledge Base is the brain of your AI employees. It's where you store all the documents, information, and content that your AI uses to answer questions accurately. The better your Knowledge Base, the smarter your AI.

## What Is the Knowledge Base?

Think of it as a library for your AI. You add documents, and your AI employees can reference them when answering questions. This could include:

* Product documentation
* FAQs and help articles
* Company policies
* Training materials
* Website content
* Guides and tutorials

Without a Knowledge Base, your AI only knows general information. With a good Knowledge Base, it becomes an expert on YOUR business.

## Adding Your First Document

### Option 1: Upload a File

1. Go to **Knowledge Base** in the sidebar
2. Click **Upload** or **Add Document**
3. Select a file from your computer
4. Wait for it to process
5. Done! Your AI can now use this information

**Supported file types:**

* PDF (.pdf)
* Word documents (.doc, .docx)
* Text files (.txt)
* Spreadsheets (.xlsx, .csv)
* PowerPoint (.pptx)
* HTML files

### Option 2: Create a Document Manually

1. Click **Create Document** or **New**
2. Give it a title
3. Type or paste your content
4. Click **Save**

This is great for quick notes, FAQs, or information not in a file.

### Option 3: Import from a Website

1. Click **Import Website** or the web crawler icon
2. Enter the URL you want to import
3. Choose how many pages to crawl (just this page, or follow links)
4. Click **Start**
5. The content is converted into documents

Perfect for importing your existing help center or website content.

## Organizing with Folders

As your Knowledge Base grows, keep it organized:

### Creating a Folder

1. Click **New Folder**
2. Name your folder (e.g., "Product Docs", "Policies", "Training")
3. Click **Create**

### Moving Documents to Folders

1. Find the document you want to move
2. Drag it onto the folder, OR
3. Click the **three-dot menu** (⋮) → **Move to Folder**
4. Select the destination folder

### Navigating Folders

* Click a folder to see its contents
* Use the breadcrumb trail at the top to go back
* Click the home icon to return to the root level

## Using Tags

Tags help you categorize documents across folders:

### Adding Tags to a Document

1. Click on a document to open it
2. Find the **Tags** field
3. Type a tag name and press Enter
4. Add as many tags as needed
5. Save

### Adding Tags to an Entire Folder

Apply tags to all documents in a folder at once:

1. Click the **three-dot menu** (⋮) on a folder
2. Select **Manage Tags**
3. Add your tags
4. Save

### Using Tags to Find Documents

* Click on a tag to filter documents with that tag
* Use the search bar with tag names

## Document Sharing Options

Control who can access each document:

| Setting     | Who Can See It           |
| ----------- | ------------------------ |
| **Private** | Only you                 |
| **Shared**  | Everyone in your company |
| **Public**  | Anyone with the link     |

### Changing Sharing Settings

1. Click on a document
2. Click **Sharing** or look for the privacy icon
3. Select your preferred setting
4. Save

## Assigning Documents to Specific Employees

Want certain AI employees to only know certain things?

1. Click on a document
2. Find **Assigned Employee** or **Access**
3. Select which employees should have access
4. Save

**Use cases:**

* Sales AI only knows sales materials
* Support AI only knows support docs
* Different products have different AI assistants

## Managing Your Documents

### Editing a Document

1. Click on the document
2. Click **Edit**
3. Make your changes
4. Click **Save**

### Deleting a Document

1. Click the **three-dot menu** (⋮) on the document
2. Select **Delete**
3. Confirm

**Note:** Deleted documents are gone permanently. Make sure you have backups of important content.

### Bulk Operations

Select multiple documents to perform bulk actions:

1. Check the boxes next to multiple documents
2. Use the bulk action menu to:
   * Delete selected
   * Move to folder
   * Apply tags
   * Change sharing

## Tips for a Great Knowledge Base

### 1. Quality Over Quantity

A few well-written, comprehensive documents are better than many scattered, incomplete ones.

### 2. Keep It Current

Outdated information leads to wrong answers. Review and update your Knowledge Base regularly.

### 3. Be Specific

Instead of "Our return policy is flexible," write "We accept returns within 30 days of purchase. Items must be unused and in original packaging."

### 4. Include Common Questions

Add an FAQ document with questions customers actually ask and the answers they need.

### 5. Use Clear Titles

Name documents clearly so both you and the AI can find them:

* **Avoid:** "doc\_v2\_final\_FINAL"
* **Better:** "Return Policy - Updated January 2024"

### 6. Structure Your Content

Use headings, bullet points, and clear sections. This helps the AI find relevant information quickly.

## How Document Processing Works

When you upload a document, here's what happens:

1. **Upload:** Your file is securely stored
2. **Extraction:** Text is pulled from the document
3. **Processing:** The content is analyzed and indexed
4. **Ready:** Your AI can now reference this information

Processing usually takes a few seconds to a few minutes depending on document size.

## Troubleshooting

### "My AI doesn't know about a document I added"

* Make sure the document finished processing (no "processing" indicator)
* Check that the document is shared with the AI employee you're using
* Verify the information is actually in the document text (not just images)

### "Upload failed"

* Check file size (there may be limits)
* Make sure the file type is supported
* Try a different browser or clear your cache

### "The AI gives old information"

* Update your documents with current information
* Remove outdated documents
* Add version dates to document titles

## Frequently Asked Questions

**Q: How much can I upload?**
A: Storage limits depend on your subscription. Check your Profile for your limits.

**Q: Can the AI read images in documents?**
A: Currently, the AI primarily reads text. Important information in images should also be included as text.

**Q: How long until a new document is usable?**
A: Usually within a few minutes. Large documents may take longer.

**Q: Can I connect to external sources like Google Drive?**
A: Yes! Check the [Integrations](/settings/integrations) page for available connections.

## Next Steps

* [Create AI Employees](/overview/employees)
* [Start Chatting](/overview/chat)
* [Connect External Tools](/settings/integrations)


# Start Here
Source: https://docs.parallellabs.app/overview/overview

The tools you use every day: your AI employees, the conversations you have with them, and the work they do on their own.

These are the items at the top of the sidebar, above the feature sections. Everything else in the platform is built on them — an AI employee with a knowledge base is what powers agents, sequences, content, and automations.

## In This Section

* **[AI Employees](/overview/employees)** — Create and customize the AI team members that power chat, agents, and automations.
* **[Chat](/overview/chat)** — Talk directly with your AI employees to get work done, ask questions, and think out loud.
* **[Scheduled](/overview/scheduled)** — Give an employee standing work that runs once, on a schedule, or continuously.
* **[Knowledge Base](/overview/knowledge-base)** — Upload the documents, websites, and files that train your AI employees.
* **[Dashboard](/overview/dashboard)** — Your analytics home: activity, usage, and performance across every feature you use.

## How These Fit Together

An **AI employee** is a persona with its own instructions, model, and access to your data. You give it knowledge through the **Knowledge Base**, work with it in **Chat**, and hand it standing work in **Scheduled**. The **Dashboard** shows what all of it added up to.

The same employees show up throughout the rest of the platform — as the brain behind an [agent](/inbound/agents), the writer behind [content](/engage/content-engine), and the researcher behind [Smart List](/find-enrich/smart-lists) actions.


# Scheduled
Source: https://docs.parallellabs.app/overview/scheduled

Give an AI employee standing work: tasks that run once, on a repeating schedule, or continuously through the working day.

Chat is work you ask for in the moment. **Scheduled** is work that happens without you — a task you write once, assigned to an AI employee, that runs on its own and reports back.

## What a Scheduled Task Is

A task is a prompt plus a schedule. At the scheduled time, the prompt is sent to the AI employee you picked, exactly as if you had typed it in chat. The employee does the work using its knowledge base, integrations, and tools, and the result is saved to the task's run history.

Typical uses:

* A morning briefing of overnight replies and new leads
* A weekly summary of campaign performance
* Recurring research — competitor pricing, industry news, brand mentions
* Routine list upkeep, such as enriching new rows
* Anything you would otherwise remember to ask for every day

## Creating a Task

1. Select your company from the sidebar
2. Click **Scheduled**
3. Click **New Task**
4. Fill in the task:
   * **Task Name** — how it appears in the list
   * **AI Employee** — who does the work
   * **Prompt** — the instruction sent at the scheduled time
5. Choose a schedule type (below)
6. Save

The prompt is the whole instruction. Write it as if the employee has no memory of the conversation, because each run starts fresh.

## Schedule Types

| Type          | When it runs                                            | Good for                                                         |
| ------------- | ------------------------------------------------------- | ---------------------------------------------------------------- |
| **One-time**  | Once, on a date and time you pick                       | A reminder or a one-off job you don't want to do now             |
| **Recurring** | Daily, weekly, or monthly at a set time                 | Briefings, reports, regular check-ins                            |
| **Heartbeat** | Repeatedly on an interval, during working hours you set | Work that should be picked up through the day, not at one moment |

### One-time

Pick a **Date** and a **Time**. The task runs once and then stops.

### Recurring

Pick a **Frequency** — Daily, Weekly, or Monthly — and a **Time**. Weekly tasks also let you tick which **Days of Week** to run on.

### Heartbeat

A heartbeat task runs continuously on an interval during the active hours you set, rather than at a single scheduled moment. Set the **Interval**, the **Start** and **End** of its active hours, and the days of the week it should be active.

This suits work that needs picking up as it arrives — watching for replies, reacting to new rows in a list — instead of being batched into one daily run.

## Managing Tasks

The task list shows each task with its **Schedule**, **Next Run**, number of **Runs**, and an **Avg Score** with a trend, plus a status dot.

From a task's menu you can:

* **Edit** — change the prompt, employee, or schedule
* **Pause** / **Resume** — stop a task running without deleting it
* **Evaluate** — score how well the runs are going
* **Reset Task History** — clear the run history, evaluations, and daily counters
* **Delete** — remove the task

Click a task to open its run history and read what the employee actually did on each run.

## Scoring and Evaluation

Runs can be scored so you can see whether a task is doing useful work over time. The **Avg Score** column and the trend next to it summarize this at a glance; open **Evaluate** on a task for the detail.

A task whose score is drifting down usually needs a sharper prompt — see the prompt tips in [Chat](/overview/chat).

## Tips

1. **Be explicit about output.** "Summarize in five bullets" produces a more consistent result than "tell me what happened."
2. **Pick the right employee.** The task inherits that employee's instructions, knowledge, and integrations.
3. **Start with one-time.** Run a task once to see the output before putting it on a schedule.
4. **Use heartbeat sparingly.** It runs many times a day and uses credits on every run.
5. **Pause instead of deleting** while you are tuning a prompt — you keep the history.

## Next Steps

* [Create an AI Employee](/overview/employees)
* [Chat with your AI employees](/overview/chat)
* [Add Knowledge for tasks to draw on](/overview/knowledge-base)


# Manage Parallel AI Company API Keys Programmatically
Source: https://docs.parallellabs.app/platform/api-keys

Create, list, retrieve, and delete API keys for your Parallel AI company. Use the /api/v0/api_keys endpoints to automate key provisioning at scale.

## Key Types

Parallel AI supports two types of API keys:

<Accordion title="Company API Keys">
  Company API Keys are scoped to a specific company account. You can find and create them on the **Integrations** page in your Parallel AI dashboard. Use these for server-side integrations where all requests belong to one company.
</Accordion>

<Accordion title="Personal Access Keys (PAKs)">
  Personal Access Keys are tied to your user account and can act on behalf of any company you belong to. Create them on your **Profile** page. Every request made with a PAK must also include an `X-Company-ID` header to identify which company context to use.
</Accordion>

## Programmatic Key Management

You can manage Company API Keys via the `/api/v0/api_keys` endpoints. This is useful for automated provisioning and multi-tenant setups.

### List Keys

`GET /api/v0/api_keys` — Retrieve all API keys associated with a company. Pass the `companyId` as a query parameter:

```bash theme={null}
curl "https://api.parallellabs.app/api/v0/api_keys?companyId=your_company_id" \
  -H "X-API-Key: YOUR_API_KEY"
```

**Response:**

```json theme={null}
{
  "apiKey": "the_existing_api_key_value"
}
```

### Create a Key

`POST /api/v0/api_keys` — Create a new API key with a descriptive name. Both `companyId` and `name` are required:

```bash theme={null}
curl -X POST "https://api.parallellabs.app/api/v0/api_keys" \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "companyId": "your_company_id",
    "name": "production-integration"
  }'
```

**Response:**

```json theme={null}
{
  "apiKey": "newly_generated_api_key_value"
}
```

<Warning>
  Copy the returned `apiKey` value immediately — it will not be displayed again after this response.
</Warning>

### Get a Specific Key

`GET /api/v0/api_keys/{companyId}/{name}` — Retrieve a single key by company ID and key name:

```bash theme={null}
curl "https://api.parallellabs.app/api/v0/api_keys/your_company_id/production-integration" \
  -H "X-API-Key: YOUR_API_KEY"
```

**Response:**

```json theme={null}
{
  "apiKey": "the_api_key_value"
}
```

Returns `404` if no key with that name exists for the given company.

### Create or Get Existing Key (Upsert)

`PUT /api/v0/api_keys` — Create a key if it doesn't exist, or return the existing one if it does. This is ideal for idempotent provisioning scripts that run on every deployment:

```bash theme={null}
curl -X PUT "https://api.parallellabs.app/api/v0/api_keys" \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "companyId": "your_company_id",
    "name": "production-integration"
  }'
```

**Response:**

```json theme={null}
{
  "apiKey": "the_api_key_value"
}
```

### Delete a Key

`DELETE /api/v0/api_keys` — Permanently delete a key by passing the company ID and key name in the request body:

```bash theme={null}
curl -X DELETE "https://api.parallellabs.app/api/v0/api_keys" \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "companyId": "your_company_id",
    "name": "production-integration"
  }'
```

<Warning>
  Deleting an API key is permanent and cannot be undone. Any integration using the deleted key will immediately stop working.
</Warning>

## Security Best Practices

<Accordion title="Use descriptive, purpose-specific names">
  Name each key after its use case (e.g. `crm-sync`, `marketing-automation`, `staging-tests`). This makes it easy to audit which systems have access and revoke individual keys without disrupting others.
</Accordion>

<Accordion title="Rotate keys regularly">
  Periodically replace active keys — especially after team member offboarding or a suspected exposure. Use the upsert endpoint to provision a new key, update your integration, then delete the old one.
</Accordion>

<Accordion title="Store keys in environment variables">
  Never hard-code API keys in your source code. Load them from environment variables or a secrets manager at runtime:

  ```python theme={null}
  import os
  api_key = os.environ["PARALLEL_AI_API_KEY"]
  ```
</Accordion>

<Accordion title="Never commit keys to source control">
  Add your `.env` file to `.gitignore` and audit your repository history if you suspect a key was accidentally committed. Treat any exposed key as compromised and rotate it immediately.
</Accordion>

## Learn More

* [Authentication](/authentication) — how to include your API key in requests
* [API Keys Reference](/api-reference/introduction) — full endpoint documentation


# Parallel AI API Error Reference and Retry Strategies
Source: https://docs.parallellabs.app/platform/errors

Parallel AI returns standard HTTP status codes with JSON error messages. Learn what each code means and how to handle errors in your integration.

## Error Response Format

Every error response from the Parallel AI API returns a JSON body with a single `error` field describing what went wrong:

```json theme={null}
{
  "error": "Description of what went wrong"
}
```

Always check both the HTTP status code and the `error` message when debugging — the message provides specific context that the status code alone cannot.

## HTTP Status Codes

| Code  | Meaning               | Common Cause                                      |
| ----- | --------------------- | ------------------------------------------------- |
| `200` | Success               | —                                                 |
| `201` | Created               | Resource successfully created                     |
| `400` | Bad Request           | Missing required field or invalid parameter value |
| `401` | Unauthorized          | Missing, invalid, or expired API key              |
| `403` | Forbidden             | Feature not available on your current plan        |
| `404` | Not Found             | Resource ID does not exist                        |
| `429` | Too Many Requests     | Rate limit exceeded                               |
| `500` | Internal Server Error | Unexpected server-side error                      |

## Common Errors and Solutions

<Accordion title="401 Unauthorized — invalid or missing API key">
  **Symptom:** You receive `{"error": "Unauthorized"}` on every request.

  **Solutions:**

  * Confirm you're sending the `X-API-Key` header (or `Authorization: Bearer ...`) with every request.
  * Verify the key is correct — copy it fresh from the dashboard to rule out typos.
  * If you're using a **Personal Access Key** (PAK), you must also include the `X-Company-ID` header:

  ```bash theme={null}
  curl "https://api.parallellabs.app/api/v0/agents" \
    -H "Authorization: Bearer pak_your_key" \
    -H "X-Company-ID: your_company_id"
  ```
</Accordion>

<Accordion title="403 Forbidden — feature not enabled on your plan">
  **Symptom:** Your request is authenticated but returns `{"error": "Forbidden"}`.

  **Solutions:**

  * The endpoint or feature you're calling may not be included in your current Parallel AI plan.
  * Contact [Parallel AI support](mailto:support@parallellabs.app) to confirm which features are available on your plan or to request access.
</Accordion>

<Accordion title="404 Not Found — resource does not exist">
  **Symptom:** A resource lookup returns `{"error": "Not found"}`.

  **Solutions:**

  * Double-check the resource ID in your request path (e.g. agent ID, sequence ID).
  * Confirm the resource belongs to the company associated with your API key — you cannot access resources from other companies.
  * The resource may have been deleted. Use a list endpoint to verify it still exists.
</Accordion>

<Accordion title="429 Too Many Requests — rate limit exceeded">
  **Symptom:** Requests start returning `{"error": "Too many requests"}`.

  **Solutions:**

  * Slow down the rate at which you're making requests.
  * Implement **exponential backoff**: wait before retrying, doubling the wait time on each successive failure.
  * Batch operations where possible, or use `pageSize=100` to retrieve more data per request.
</Accordion>

<Accordion title="500 Internal Server Error">
  **Symptom:** The API returns a `500` status unexpectedly.

  **Solutions:**

  * This indicates an issue on Parallel AI's servers, not your request.
  * Retry the request after a short delay — most transient errors resolve quickly.
  * If the error persists, contact [Parallel AI support](mailto:support@parallellabs.app) with the request details and timestamp.
</Accordion>

## Retry Logic with Exponential Backoff

For production integrations, implement retry logic to handle transient errors (`429` and `500`) gracefully:

```python theme={null}
import requests
import time

def api_call_with_retry(url, headers, max_retries=3):
    for attempt in range(max_retries):
        response = requests.get(url, headers=headers)

        if response.status_code == 429:
            wait = 2 ** attempt  # 1s, 2s, 4s
            print(f"Rate limited. Retrying in {wait}s...")
            time.sleep(wait)
            continue

        response.raise_for_status()
        return response.json()

    raise Exception("Max retries exceeded")
```

<Tip>
  Add jitter (a small random delay) to your backoff to prevent multiple clients from retrying simultaneously and creating a thundering-herd effect:

  ```python theme={null}
  import random
  wait = (2 ** attempt) + random.uniform(0, 1)
  time.sleep(wait)
  ```
</Tip>

<Warning>
  Do not retry `400`, `401`, `403`, or `404` errors — these indicate problems with your request or credentials that will not resolve on their own.
</Warning>


# Paginate Parallel AI List Endpoint Responses Easily
Source: https://docs.parallellabs.app/platform/pagination

All Parallel AI list endpoints support cursor-free pagination using page and pageSize query parameters. Learn how to iterate through large result sets.

## Overview

Every list endpoint in the Parallel AI API uses **page-based pagination**. You control which page of results to return and how many items appear per page using two query parameters:

| Parameter  | Default | Maximum | Description                  |
| ---------- | ------- | ------- | ---------------------------- |
| `page`     | `1`     | —       | The page number to retrieve  |
| `pageSize` | `10`    | `100`   | The number of items per page |

## Paginated Response Format

All paginated responses include the result items alongside metadata to help you navigate through the full dataset:

| Field     | Type    | Description                                              |
| --------- | ------- | -------------------------------------------------------- |
| `items`   | array   | The list of results for the current page                 |
| `total`   | integer | The total number of items across all pages               |
| `pages`   | integer | The total number of pages at the current `pageSize`      |
| `hasNext` | boolean | `true` if there is at least one more page after this one |

## Example Request

Fetch the second page of agents, with 20 items per page:

```bash theme={null}
curl "https://api.parallellabs.app/api/v0/agents?page=2&pageSize=20" \
  -H "X-API-Key: YOUR_API_KEY"
```

## Example Response

```json theme={null}
{
  "items": [
    { "id": "agent_021", "name": "Outreach Agent", "status": "active" },
    { "id": "agent_022", "name": "Research Agent", "status": "active" }
  ],
  "total": 57,
  "pages": 3,
  "hasNext": true
}
```

<Note>
  Use `hasNext` rather than comparing `page` to `pages` to determine whether to continue — it accounts for edge cases where the total changes between requests.
</Note>

## Fetching All Pages

To retrieve all items across every page, loop until `hasNext` is `false`. The example below uses `pageSize=100` to minimise the number of requests:

```python theme={null}
import requests

def get_all_agents(api_key):
    agents = []
    page = 1
    while True:
        resp = requests.get(
            "https://api.parallellabs.app/api/v0/agents",
            headers={"X-API-Key": api_key},
            params={"page": page, "pageSize": 100}
        )
        data = resp.json()
        agents.extend(data["items"])
        if not data["hasNext"]:
            break
        page += 1
    return agents
```

<Tip>
  Set `pageSize` to `100` (the maximum) when fetching all pages to minimise the total number of API requests and reduce latency.
</Tip>

## Tips for Large Datasets

<Accordion title="Filter before paginating">
  Use any available filter parameters (e.g. `status`, `search`, date ranges) to narrow the result set before paginating. Fewer total items means fewer pages to traverse and faster response times.
</Accordion>

<Accordion title="Handle rate limits gracefully">
  If you're iterating through many pages in a tight loop, add a short delay between requests and implement retry logic for `429 Too Many Requests` responses. See [Errors](/platform/errors) for a retry example.
</Accordion>

<Accordion title="Cache total counts">
  The `total` and `pages` fields are useful for displaying progress indicators or estimating job duration. Cache them from the first response rather than re-fetching metadata on every page.
</Accordion>


# Quickstart Guide: Make Your First Parallel AI API Call
Source: https://docs.parallellabs.app/quickstart

Learn how to authenticate with Parallel AI and make your first API call to list agents, create a chat completion, and verify everything is working.

## Prerequisites

All you need is a Parallel AI account and a terminal (or any HTTP client). No SDK installation is required to follow this guide — though we'll show you a Python example too.

<Steps>
  <Step title="Create your Parallel AI account">
    Head to [https://web.parallellabs.app/signup](https://web.parallellabs.app/signup) and sign up using your Google account or email address.

    Once you're in, you'll land on your company dashboard. If this is your first time, a default company workspace is created for you automatically.
  </Step>

  <Step title="Generate a Company API Key">
    1. In your dashboard, navigate to the **Integrations** page from the left sidebar.
    2. Find the **API Keys** section and click **Generate New Key**.
    3. Give your key a descriptive name (e.g. `my-first-key`).
    4. Copy the key immediately — it **won't be shown again** after you close the dialog.

    <Warning>
      Store your API key somewhere safe, such as a password manager or secrets vault. If you lose it, you'll need to generate a new one.
    </Warning>

    You'll pass this key in the `X-API-Key` header on every request. See the [Authentication guide](/authentication) for all supported header formats.
  </Step>

  <Step title="List your agents">
    With your key in hand, make your first live API call. This request retrieves all agents in your company:

    ```bash theme={null}
    curl -X GET "https://api.parallellabs.app/api/v0/agents" \
      -H "X-API-Key: YOUR_API_KEY" \
      -H "Content-Type: application/json"
    ```

    If your account is brand new, you'll see an empty list — that's expected:

    ```json theme={null}
    {
      "items": [],
      "total": 0,
      "pages": 1,
      "hasNext": false
    }
    ```

    A `200 OK` response confirms that your key is valid and your connection to the API is working correctly.

    <Tip>
      Replace `YOUR_API_KEY` with the key you copied in the previous step. Keep this pattern in mind — every request to the Parallel AI API uses the same `X-API-Key` header.
    </Tip>
  </Step>

  <Step title="Create your first chat completion">
    Parallel AI exposes an OpenAI-compatible `/chat/completions` endpoint. Use it to run a completion against any supported model:

    <Tabs>
      <Tab title="cURL">
        ```bash theme={null}
        curl -X POST "https://api.parallellabs.app/api/v0/chat/completions" \
          -H "X-API-Key: YOUR_API_KEY" \
          -H "Content-Type: application/json" \
          -d '{
            "model": "gpt-4o-mini",
            "messages": [{"role": "user", "content": "Hello!"}]
          }'
        ```
      </Tab>

      <Tab title="Python (openai library)">
        Because Parallel AI's chat completions endpoint is OpenAI-compatible, you can use the official `openai` Python package with no changes other than the `base_url`:

        ```python theme={null}
        from openai import OpenAI

        client = OpenAI(
            api_key="YOUR_API_KEY",
            base_url="https://api.parallellabs.app/api/v0",
        )

        response = client.chat.completions.create(
            model="gpt-4o-mini",
            messages=[{"role": "user", "content": "Hello!"}],
        )

        print(response.choices[0].message.content)
        ```

        Install the library first if you haven't already:

        ```bash theme={null}
        pip install openai
        ```
      </Tab>
    </Tabs>

    The API returns a standard OpenAI-format response object with the model's reply inside `choices[0].message.content`.

    <Note>
      You can swap `"gpt-4o-mini"` for any model supported on the platform. Refer to the [API Reference](/api-reference/introduction) for the full list of available models.
    </Note>
  </Step>

  <Step title="You're ready — explore what's next">
    You've authenticated, listed your agents, and run a live chat completion. Here's where to go from here:

    <CardGroup>
      <Card title="Build AI Agents" icon="robot" href="/inbound/agents">
        Create multi-channel agents for web, email, SMS, and voice — and connect them to your Knowledge Base.
      </Card>

      <Card title="Full API Reference" icon="code" href="/api-reference/introduction">
        Explore every endpoint available in the Parallel AI REST API with request and response schemas.
      </Card>
    </CardGroup>
  </Step>
</Steps>


# Collaborators
Source: https://docs.parallellabs.app/settings/collaborators

Invite team members, assign roles, and organize access by department.

Add team members to your company so they can help manage your AI employees, content, and operations. Control who has access and what they can do.

## What Are Collaborators?

Collaborators are team members who have access to your company. Instead of sharing login credentials, you can invite people to join and assign them specific roles.

### Why Add Collaborators?

* **Teamwork:** Multiple people can manage your AI employees
* **Security:** Each person has their own login
* **Control:** Assign roles to limit what people can access
* **Accountability:** Track who made changes

## Adding a New Team Member

### Step 1: Open Collaborators

1. Select your company from the sidebar
2. Click **Collaborators** (under Settings)
3. You'll see a list of current team members

### Step 2: Invite Someone

1. Click **Invite Collaborator** (or the **+** button)
2. Enter their email address
3. Select a role for them
4. Click **Send Invite**

### Step 3: They Accept

1. The person receives an email invitation
2. They click the link to accept
3. If they don't have an account, they create one
4. Once accepted, they appear in your collaborators list

## Understanding Roles

Different roles have different permissions:

### Owner

Complete control over the company:

* Access to everything
* Can delete the company
* Can manage all users
* Can configure all settings

**Note:** Each company has one owner. Ownership can be transferred.

### Admin

Full access to features and users:

* Manage AI employees
* Access all features
* Invite and remove collaborators
* Change company settings

**Cannot:** Delete the company

**Best for:** Business managers, department heads, trusted team leads

### Member

Standard access for day-to-day work:

* Use AI employees
* Create and manage content
* View dashboards
* Access most features

**Cannot:** Manage other users, change critical settings

**Best for:** Team members, content creators, regular users

> Want to give someone a narrower view? Keep them as a **Member** and use **Manage Features** (below) to turn off the features they don't need.

## Managing Existing Collaborators

### Viewing Your Team

1. Go to **Collaborators**
2. See all team members with their:
   * Name and email
   * Role
   * Departments (if configured)
   * Status (active, pending)

### Changing Someone's Role

1. Find the person in the list
2. Click the **three-dot menu** (⋮)
3. Select **Change Role**
4. Choose the new role
5. Confirm

### Managing Feature Access

Control which features each user can access:

1. Find the person in the list
2. Click **Manage Features**
3. Enable or disable specific features:
   * Chat
   * Knowledge Base
   * Sequences
   * Smart Lists
   * Content Engine
   * Images, Audio, Video
   * Integrations
4. Save changes

### Removing a Collaborator

1. Find the person in the list
2. Click the **three-dot menu** (⋮)
3. Select **Remove**
4. Confirm the removal

**Note:** Removing someone immediately revokes their access to the company.

## Organizing with Departments

Group users by department for better organization.

### Creating Departments

1. Click **Departments** button
2. Click **Add Department**
3. Enter department name
4. Save

### Assigning Users to Departments

1. Find the user
2. Click **Manage Departments**
3. Select their department(s)
4. Save

### Why Use Departments?

* Organize large teams
* Restrict content by department
* Track usage by group
* Simplify reporting

## Pending Invitations

When you invite someone, they appear as "Pending" until they accept.

### Checking Invitation Status

Look for the status indicator:

* **Pending:** Waiting for them to accept
* **Active:** They've accepted and have access

### Resending an Invitation

If someone didn't receive or lost the email:

1. Find them in the list
2. Click the **three-dot menu** (⋮)
3. Select **Resend Invite**

### Canceling an Invitation

Changed your mind? Cancel before they accept:

1. Find the pending invitation
2. Click the **three-dot menu** (⋮)
3. Select **Remove** — this cancels the pending invite

## Single Sign-On (SSO)

For enterprise teams, set up SSO for easier login management.

### Supported Providers

* Okta
* Azure Active Directory
* Google Workspace
* OneLogin
* Any SAML 2.0 provider

### Setting Up SSO

1. Click **SSO Settings**
2. Follow the setup wizard
3. Choose your identity provider
4. Configure SAML settings
5. Exchange metadata with your provider
6. Test the connection
7. Enable for users

**Note:** SSO setup typically requires IT administrator involvement.

## Best Practices for Team Management

### 1. Use the Right Role

Don't give everyone Admin access. Match roles to responsibilities:

* Only company owners need Owner access
* Managers work well as Admins
* Most team members work fine as Members
* For limited access, keep someone a Member and turn off features they don't need

### 2. Review Access Regularly

Check your collaborators list periodically:

* Remove people who left the team
* Update roles if responsibilities changed
* Clean up old pending invitations

### 3. Use Feature Access for Fine-Tuning

If someone needs a role but not all features:

1. Assign their base role
2. Disable features they don't need
3. This provides better security

### 4. Document Who Does What

For larger teams, keep notes on:

* Who manages which AI employees
* Who creates content
* Who handles integrations

## Subscription and Seats

Your subscription determines how many collaborators you can have.

### Checking Your Limits

1. Go to your [Profile](/account/profile)
2. View your subscription details
3. See how many seats you have vs. how many are used

### Adding More Seats

If you need more team members than your plan allows:

1. Go to your Profile
2. Click **Manage Subscription**
3. Upgrade your plan or add seats

## Troubleshooting

### "Invitation not received"

* Check their spam/junk folder
* Verify the email address is correct
* Try resending the invitation
* Have them check if they already have an account

### "Can't accept invitation"

* The invitation might have expired (resend it)
* They might already be a collaborator
* They need to create an account first

### "Can't remove a collaborator"

* You need Admin or Owner role to remove others
* You cannot remove yourself if you're the owner
* There must be at least one owner

### "Access denied" for a collaborator

* Check their role has the right permissions
* Verify their feature access settings
* Make sure they're looking at the correct company
* Confirm their account is active

## Frequently Asked Questions

**Q: Can I be a collaborator on multiple companies?**
A: Yes! You can be invited to as many companies as needed, each with its own role.

**Q: What happens when I remove someone?**
A: They immediately lose access. Their past actions remain in the system logs.

**Q: Can collaborators see billing information?**
A: Only Owners can see billing and subscription details.

**Q: Can I transfer ownership of a company?**
A: Yes, see [Companies](/account/companies) for how to transfer ownership.

**Q: Do collaborators need their own subscription?**
A: No, they use seats on the company's subscription.

**Q: Can I restrict someone to specific features only?**
A: Yes, use the Manage Features option to control feature access per user.

## Next Steps

* [Set Up Integrations](/settings/integrations)
* [Create AI Employees](/overview/employees)
* [Manage Your Companies](/account/companies)


# Integrations
Source: https://docs.parallellabs.app/settings/integrations

Connect CRMs, email, data providers, social platforms, and other services.

Connect your AI employees to the tools you already use. Integrations let your AI access email, CRM, social media, and other services to work more effectively.

## What Are Integrations?

Integrations connect your AI Employee platform to external services. Once connected, your AI employees can:

* Read and send emails
* Access your CRM contacts
* Post to social media
* Pull data from other tools
* Automate across platforms

## Available Integrations

The exact list you see depends on your plan and region. Common integrations include:

### Communication

| Integration                             | What It Does                          |
| --------------------------------------- | ------------------------------------- |
| **Slack**                               | Send messages and respond in channels |
| **Microsoft Teams**                     | Team messaging                        |
| **Discord**                             | Bot responses in Discord servers      |
| **Prosp**, **HeyReach**, **Salesforge** | Push contacts to outreach campaigns   |

### Email & Sending

| Integration               | What It Does                           |
| ------------------------- | -------------------------------------- |
| **Gmail**                 | Send and receive with a Google account |
| **Office 365 Email**      | Send and receive with Microsoft 365    |
| **IMAP Email**            | Connect any IMAP/SMTP mailbox          |
| **AWS SES**, **SendGrid** | High-volume transactional sending      |
| **Twilio**                | Send SMS, voice, and WhatsApp messages |

### CRM & Data Enrichment

| Integration        | What It Does                           |
| ------------------ | -------------------------------------- |
| **HubSpot**        | Sync contacts, deals, and companies    |
| **Salesforce**     | CRM data and automation                |
| **GoHighLevel**    | CRM and marketing automation           |
| **Apollo**         | Enrich and reveal contact data         |
| **Findymail**      | Find and verify emails                 |
| **Better Contact** | Enrich contacts with emails and phones |

### Social & Advertising

| Integration                 | What It Does                 |
| --------------------------- | ---------------------------- |
| **LinkedIn**                | Profile access and messaging |
| **Facebook**, **Instagram** | Page and profile access      |
| **Meta Ads**                | Build and sync ad audiences  |
| **Google Analytics**        | Pull site analytics          |

### Content & Knowledge

| Integration                | What It Does                          |
| -------------------------- | ------------------------------------- |
| **Google Drive**           | Sync documents to your Knowledge Base |
| **Notion**, **Confluence** | Sync workspace pages                  |
| **Website Crawler**        | Import website content                |
| **WordPress**, **Webflow** | Publish content to your site          |

### Other

| Integration             | What It Does                                |
| ----------------------- | ------------------------------------------- |
| **Shopify**             | Connect your store                          |
| **Calendar Scheduling** | Let contacts book meetings                  |
| **Thanks.io**           | Send physical direct mail                   |
| **HeyGen**              | Generate AI video                           |
| **Browser Session**     | Power browser-automation tasks              |
| **Zapier**              | Connect to thousands of other apps          |
| **MCP Tools**           | Connect Model Context Protocol tool servers |

## Setting Up Your First Integration

### Step 1: Open Integrations

1. Select your company from the sidebar
2. Click **Integrations** (under Settings)
3. Browse available integrations

### Step 2: Choose an Integration

1. Find the service you want to connect
2. Click on it to see details
3. Click **Connect** or **Install**

### Step 3: Authorize Access

Most integrations use OAuth (secure authorization):

1. A new window opens for the service
2. Log in to your account (if needed)
3. Review what permissions are requested
4. Click **Allow** or **Authorize**
5. The window closes and you're connected

### Step 4: Configure Settings

Some integrations have additional settings:

* Which account to use (if you have multiple)
* What data to access
* Specific permissions to enable

## Managing Connected Integrations

### Viewing Your Integrations

1. Go to **Integrations**
2. See all connected services
3. Status shows if each is active or needs attention

### Checking Connection Status

Look for status indicators:

| Status                    | What It Means                   |
| ------------------------- | ------------------------------- |
| **Connected**             | Working normally                |
| **Needs Reauthorization** | Token expired, reconnect needed |
| **Error**                 | Something went wrong            |
| **Syncing**               | Data sync in progress           |

### Reconnecting an Integration

If a connection breaks (tokens expire, password changed):

1. Find the integration
2. Click **Reconnect** or **Reauthorize**
3. Complete the authorization flow again

### Removing an Integration

1. Find the integration
2. Click the **three-dot menu** (⋮) or **Remove**
3. Confirm removal

**Note:** Removing an integration stops AI employees from accessing that service.

## API Keys

Create API keys to connect external systems to your AI Employee platform.

### Creating an API Key

1. Go to **Integrations**
2. Find the **API Keys** section
3. Click **Create Key**
4. Name your key (e.g., "Website Integration")
5. Click **Create**
6. **Copy the key immediately** — you won't see it again!

An API key is scoped to the company it was created in and acts on behalf of that company. Create separate keys for separate systems so you can revoke one without affecting the others.

### Using Your API Key

Include the key in your API requests:

```
Authorization: Bearer YOUR_API_KEY
```

See the full API reference from the **API Documentation** link in the documentation sidebar.

### Deleting an API Key

1. Find the key in the list
2. Click **Delete**
3. Confirm

**Warning:** Deleting a key immediately breaks any integrations using it.

## Integration-Specific Setup Guides

### Slack Integration

**What you can do:**

* Send messages to channels
* Respond to direct messages
* Create automated notifications

**Setup:**

1. Click **Add to Slack**
2. Select your workspace
3. Choose which channels the bot can access
4. Authorize

**Tips:**

* Create a dedicated channel for AI notifications
* Configure which events trigger responses

### HubSpot Integration

**What you can do:**

* Read contact information
* Update deal stages
* Log activities
* Sync data with Smart Lists

**Setup:**

1. Click **Connect HubSpot**
2. Log in to your HubSpot account
3. Select the portal to connect
4. Choose which objects to access (contacts, deals, companies)

### Google Drive Integration

**What you can do:**

* Import documents to your Knowledge Base
* Keep documents synced automatically
* Access shared drive content

**Setup:**

1. Click **Connect Google Drive**
2. Sign in with your Google account
3. Allow the requested permissions
4. Select folders to sync

### Website Crawler

**What you can do:**

* Import content from any website
* Add web content to Knowledge Base
* Keep content updated

**Setup:**

1. Click **Website Crawler**
2. Enter the website URL
3. Configure crawl depth (how many pages)
4. Start the import

## Using Integrations with AI Employees

Once connected, your AI employees can use integrations automatically.

### In Smart Lists

List actions use integrations to enrich and sync your data — for example, **Apollo Enrichment**, **Sync to CRM**, or an **HTTP Request** to your own API. See [Smart Lists](/find-enrich/smart-lists) for details.

### In Sequences

Email sequences use integrations to:

* Send personalized emails
* Track opens and clicks
* Update contact records

See [Sequences](/engage/sequences) for details.

### In Chat

Some integrations allow AI employees to:

* Look up information during conversations
* Take actions based on chat requests
* Access external data to answer questions

## Calling External APIs

To send data to your own systems or third-party services, use an **HTTP Request** step:

* In a **Smart List**, add the **HTTP Request** action to call any API for each row
* In a **Sequence**, add an **HTTP Request** step to fire a call at a point in the flow
* Use **Zapier** to connect to thousands of apps without writing code

These let you push events (new members, enrichment results, replies) to external endpoints, webhooks, or automation tools like Zapier and Make.

## Security and Permissions

### What Access Is Granted?

Each integration requests specific permissions. Review them before authorizing:

* **Read access:** Can view data but not change it
* **Write access:** Can create and modify data
* **Full access:** Complete control (be careful!)

### Who Can Manage Integrations?

* Only Admins and Owners can install or remove integrations
* All team members can use integrations in their work
* API keys can be restricted by permission level

### Revoking Access from the Service

You can always revoke access from the connected service itself:

1. Go to that service's settings (Google, LinkedIn, etc.)
2. Find "Connected apps" or "Third-party access"
3. Remove the AI Employee platform

## Troubleshooting Integrations

### "Connection failed"

* Try again - temporary network issues happen
* Check if the service is experiencing outages
* Make sure you're logging into the correct account

### "Authorization expired"

* Reconnect the integration
* You may need to re-enter your password
* Check if you changed your password on that service

### "Integration not working"

* Verify the connection is still active
* Check if your account on that service is in good standing
* Review if permissions were revoked

### "Data not syncing"

* Sync can take a few minutes
* Check if there are errors in the integration status
* Try disconnecting and reconnecting

### "Rate limit reached"

* Too many requests were made too quickly
* Wait an hour and try again
* Consider spreading out automated activities

## Best Practices

### 1. Start with Essential Integrations

Don't connect everything at once. Start with:

* Your primary email
* Your main CRM
* Your most-used communication tool

### 2. Use Dedicated Accounts When Possible

For heavy automation, consider:

* A dedicated email address for AI activities
* Separate social accounts for automation
* This keeps your personal accounts clean

### 3. Monitor Activity

Regularly check:

* What actions your AI employees are taking
* If integrations are working correctly
* Any error messages or warnings

### 4. Keep Credentials Secure

* Don't share API keys
* Regularly review connected apps
* Remove integrations you no longer use

### 5. Test Before Going Live

* Test integrations with sample data first
* Verify actions work as expected
* Check that data flows correctly

## Frequently Asked Questions

**Q: Are my passwords stored?**
A: No. We use OAuth which means you authorize access without sharing your password.

**Q: Can I connect multiple accounts of the same service?**
A: Usually yes. For example, you can connect multiple email accounts.

**Q: What happens if I change my password on a connected service?**
A: You may need to reconnect the integration.

**Q: Do integrations count against my subscription?**
A: Most integrations are included. Some premium integrations may have limits.

**Q: Can I limit what an integration can access?**
A: Permissions are set by the service. You can only grant or deny the requested access level.

**Q: How do I know if an integration is working?**
A: Check the status indicator on the Integrations page. Green/Connected means it's working.

## Next Steps

* [Create AI Employees](/overview/employees)
* [Build Smart Lists](/find-enrich/smart-lists)
* [Set Up Email Sequences](/engage/sequences)
* [Manage Your Knowledge Base](/overview/knowledge-base)


# Access & Connections
Source: https://docs.parallellabs.app/settings/overview

Team access and the external services your AI employees work with.

## In This Section

* **[Integrations](/settings/integrations)** — Connect CRMs, email, data providers, social platforms, and other services.
* **[Collaborators](/settings/collaborators)** — Invite team members, assign roles, and organize access by department.

## How These Relate

**Integrations** decide what your AI employees can reach. **Collaborators** decide who can use them. Most setup problems are one of the two: an employee that can't do something usually needs an integration; a person who can't see something usually needs a role change.

Both are company-scoped — connecting a CRM in one company doesn't connect it in another.