Before letting your customers interact with your AI agent, you can generate a Preview Session to speak with it directly. This uses a temporary real-time WebRTC session connected to your agent's LLM and selected voice model.
1. Generate a Preview Session
To start a test session, make a request to the /vozzo-react-web-agent-session endpoint with your agent's UUID.
Code
JavaScript
Response
You will receive securely generated WebSocket URLs and tokens for the session. These credentials expire automatically after a short period.
Code
2. Using the Preview URL
Typically, you will use the returned ws_url and token to connect via WebRTC in your frontend application.
If you just want a quick standalone UI test without writing frontend code, you can open the temporary test viewer generated in your Vozzo Platform Dashboard under your agent's Test tab.
Real-Time WebRTC Audio
Vozzo leverages real-time WebRTC infrastructure to minimize latency and ensure crystal-clear audio during browser-based agent interactions.

