Introducing Confident Monitor

WebRTC for Voice Connections

Join your voice agent through LiveKit, Pipecat, or your own signaling URL.

Included on the Enterprise plan. Book a demo, opens in a new tab. Included on the Team plan. Included on the Starter plan. Not included on the Free plan.

Overview

Use the WebRTC transport when your voice agent talks over WebRTC instead of a phone line. Confident AI joins your agent as a participant and talks to it over audio. Choose your Provider at the top of the WebRTC tab:

  • LiveKit (default): your agent runs on LiveKit.
  • Pipecat: your agent runs on Pipecat, on Pipecat Cloud or your own server.
  • Custom signaling: your agent accepts an SDP offer at a URL you host.

WebRTC calls don't ring, so the Ring timeout call limit doesn't apply and the Test call console shows Connect instead of Dial.

Configuring Your Provider

Confident AI creates a room in your LiveKit project, dispatches your agent into it, and joins as the caller.

FieldDescription
LiveKit URLYour LiveKit server URL. It must start with wss://, for example wss://your-project.livekit.cloud.
API keyThe API key of the LiveKit project your agent runs in. Found under Settings → Keys in LiveKit Cloud.
API secretThe API secret paired with the key. Used to mint the token the simulated caller joins the room with.
Agent name (optional)The agent to dispatch into the room. Leave it blank if your agent joins every new room automatically.

Testing Your Connection

Click Connect in the Test call console. The console shows Connecting… while Confident AI joins your agent. See Testing Your Connection for what each result means.

Next Steps

Scaling beyond prototype?For teams evaluating Confident AI in productionTalk to us

Last updated on

Built byConfident AI