Skip to main content
Configure API requests that run before your agent greets the caller. Use them to fetch CRM data, meeting details, or any other context you want available from the very first sentence.

​
Overview

  • When it runs: Before the agent speaks.
  • Execution order: Multiple pre-call actions run sequentially in creation order.
  • Variable scope: Variables created here are available through the entire interaction. Only variable values assigned in pre-call action can be used to personalize the Initial Message and Guidelines.
Only variables values assigned in Pre-Call Actions are available to Initial Message and Guidelines.

​
Create a Pre-Call Action

  1. Open Studio → Build → AI Agents → Actions tab → Pre-Call.
  2. Select “Create New Action” and fill in:
    • Name: Short, descriptive identifier
    • Description: What the action fetches and why
  3. Configure the request:
    • Method: GET, POST, PUT, DELETE, PATCH
    • URL: Static URL
    • Query Params: Key–value pairs, variable-aware
    • Headers: Static or variable-aware (e.g. auth tokens)
    • Body: JSON request body for non-GET methods
    • Auth Profile: Optional reusable credentials profile
  4. Click Send to test the request and preview the response (Data and Meta tabs).
  5. Map response fields to variables:
    • Pick response paths to save into variables (e.g. firstName, accountStatus).
    • These variables can be referenced as {{firstName}} in Initial Message, Guidelines, and later actions.
  6. Save.
Response previews help you build a sample structure so you can reliably map fields to variables.

​
Best Practices

  • Keep the number of pre-call requests minimal to reduce first-utterance latency.
  • Prefer returning only the fields you need from your backend to simplify mapping.
  • Use consistent, readable variable names (e.g. customerFirstName, orderId).