The Vonage Voice API can send a call's raw audio to a WebSocket endpoint you control. That gives developers a clean place to add Resemble Detect: forward the audio, get a result back every few seconds, and use it inside the same application that's handling the call. It suits teams running their own IVR, notification lines, or AI agents on Vonage.
Use an NCCO connect action to stream call audio to your server, then on to Detect's streaming API.
Transfer the call, play a verification prompt, or end it based on the result.
Analyze inbound call audio with Resemble Detect to surface AI-cloned voices before agents answer.
Resemble Identity remembers voices tied to past fraud and flags them when they return.
Zero retention mode deletes the audio after detection, which helps teams handling health or financial calls.
