For AI agents: a documentation index is available at the root level at /llms.txt and /llms-full.txt. Append /llms.txt to any URL for a page-level index, or .md for the markdown version of any page.
The Agents API provides a comprehensive interface for creating and managing voice AI agents on the Resemble AI platform. Voice agents combine automatic speech recognition (ASR), text-to-speech (TTS), large language models (LLM), and turn-taking capabilities to power interactive voice conversations.
What are Voice Agents?
Voice agents are AI-powered conversational systems that can:
Listen: Use ASR to transcribe user speech in real-time
Think: Process conversations using LLMs with custom prompts and tools
Speak: Generate natural responses using Resemble’s TTS voices
Take Turns: Manage conversation flow with intelligent turn-taking
Key Features
Configurable Components
Each agent consists of four main components:
ASR Configuration: Speech recognition settings including provider, model, and audio format
TTS Configuration: Voice selection for speech synthesis
LLM Configuration: Language model settings including prompt, provider, model, and temperature
Turn Configuration: Conversation flow settings including timeouts and detection modes
Dynamic Variables
Agents support dynamic variables for runtime context:
Populate variables via pre-call webhooks
Reference variables in prompts and tools using {{variable_name}} syntax
Update variables during conversations using tool assignments
Maximum 64 characters for keys, 2048 characters for values
Extensibility
Extend agent capabilities with:
Webhook Tools: Call external APIs during conversations
Client Tools: Trigger client-side actions
System Tools: Use platform-provided functionality (end call, transfer, voicemail detection)
Knowledge Base: Enable RAG (Retrieval Augmented Generation) for context-aware responses
Pre-Call Webhooks: Fetch data before conversations start
Post-Call Webhooks: Log analytics after conversations end
