How can I capture custom lead data from conversations?

The Lead Extraction section lets you define custom data fields that your AI agent will extract from conversations and store in lead profiles. This gives you more control over what information you collect and how your leads are structured.

This gives you full control over:

What information you collect from users
How leads are structured
What data is available for analysis and follow-up

How it works

You define custom fields (e.g., Budget, Timeline, Interest)
AI listens to conversations and identifies relevant information
When detected, the value is automatically added to the lead profile

This enhances the Leads section by making it tailored to your business needs.

How to access Custom Leads

Open your AI agent configuration
Navigate to Tools then Lead Extraction
Configure fields for that specific agent

⚠️ Note: Custom fields apply to the AI agent you are currently editing.

Adding Custom Fields

Quick Add (Recommended)

You can quickly add common fields:

Budget Range
Timeline
Preferred Contact Method
Product Interest

Simply click any option to add it instantly.

Creating a Custom Field

Click “+ Add Field”
Fill in the following:

Field Name

Internal identifier (used in API and system)
Keep it short and clear

Display Label

How the field appears in the Leads section
Example: "Budget Range", "Preferred Plan"

What to Extract

Describe what the AI should look for in the conversation
Example:
- "What is the user's budget?"
- "Which product are they interested in?"

Click Save/Create

Timezone Setting

You can configure how date-related data is interpreted:

Default: Auto-detect (browser timezone)

Where This Data Appears

Once configured:

Extracted values appear in the Lead Details view
Fields are populated automatically when detected in conversations

Tips & Best Practices

Be specific in the “What to extract” field for better accuracy
Start with a few key fields and expand over time
Use clear and consistent naming for fields
Align fields with your sales or support workflow

Common Mistakes to Avoid

Creating too many fields at once
Writing vague extraction instructions
Not reviewing extracted data in Leads
Forgetting that fields are agent-specific