Skip to main content

Overview

Dynamic configuration allows you to personalize the behavior of your voice agent during the ringing of a call. This feature offers complete flexibility to adapt the agent to the specifics of each interaction.
This configuration is intended for use during pre-call workflows.

Configuration Structure

The configuration revolves around a structured JSON object that defines all of the agent’s operational parameters:

Configuration Detail

initialMessage
string
Greeting message that the agent will use to start the conversation with the caller.
prompt
string
required
Main instructions defining the behavior, personality, and guidelines the agent should follow during the call.
synthesizer
object
required
Configuration for the agent’s voice synthesis engine.
callDurationLimit
number
default:"1800"
Maximum allowed duration for the call before hanging up, expressed in seconds.
rag
object
Configuration of the RAG (Retrieval-Augmented Generation) database.
extractors
array
List of extractors of information that the agent should use during the call.
Example
classifiers
array
List of classifiers to automatically categorize calls.
Example
actions
array
List of actions the agent can perform during the call.
recordAudio
boolean
Enables (true) or disables (false) audio call recording.

Concise Example: Changing Voice and Language

Complete Example