Two-Way Conversational Speech Assistant

A voice assistant template that enables natural two-way conversations over phone calls using websocket connections to orchestrate real-time audio streaming between callers and AI. Demonstrates interrupt handling and AI preemption capabilities for smooth conversational experiences.

No demo yetView source Fork template

The numbers

latency—

cost / min—

frameworkCustom

The stack

telephonyTwilio Voice

speech-to-textOpenAI Whisper

llmGPT-4o

text-to-speechOpenAI TTS-1

System prompt

No prompt published.

Config

config.json

{
  "voice_id": null,
  "temperature": null,
  "barge_in": true,
  "interruption_threshold_ms": null,
  "tools": [],
  "ai_speaks_first": false,
  "interrupt_handling": "input_audio_buffer.speech_started triggers Twilio Media Streams buffer clear and OpenAI conversation.item.truncate"
}

Voice AI recipes, picks, and analysis.

Get the useful new templates plus the occasional teardown of what’s working in production voice AI.