Self-hosted AI voice agent for Asterisk. Streaming STT to LLM to TTS over AudioSocket, with barge-in and tool calling.
-
Updated
Nov 12, 2026 - Python
Self-hosted AI voice agent for Asterisk. Streaming STT to LLM to TTS over AudioSocket, with barge-in and tool calling.
Dependency-free Asterisk AudioSocket protocol implementation for Node.js. TypeScript, framed slin16 codec, paced playback, barge-in.
Paced, streaming Piper TTS for real-time voice pipelines - Asterisk AudioSocket, RTP or WebSocket. Optional ElevenLabs with local fallback.
EMS (External Media Server)
A sample .Net implementation of Asterisk AudioSocket
Measure what a caller actually hears from a telephony voice agent: frame pacing, continuity and barge-in latency over AudioSocket.
Your universal telephony gateway. Inspired by Caddy and websocketd.
Design, validate, and export Asterisk call flows visually. Build IVRs, queues, routing logic, recordings, and integrations on a drag-and-drop canvas, then generate ready-to-deploy Asterisk dialplans automatically.
AI predictive dialer with 20-bot AVR pool, AudioSocket streaming, auto-disposition via LLM, and a React control panel. FastAPI + Asterisk + Pipecat + Deepgram/Groq/Edge TTS.
Official Python SDK & GitHub Action for The Rights Foundry Music Licensing API — search 85,000+ cleared human-made tracks, stream signed previews, and license WAV sync masters in one call (zero dependencies).
Put an OpenAI Realtime voice agent on your phone line. Node.js bridge for Yeastar, Asterisk (ARI or AudioSocket) and Twilio with barge-in, RTP pacing, caller lookup, live transfer to staff, webhook/e-mail/Chatwoot escalation and call-recording transcripts. Pluggable prompt, knowledge and tools.
Asterisk examples for real-time speech transcription, AudioSocket, ARI ExternalMedia, SIP.js WebRTC softphones, and Faster-Whisper integrations.
Portable protocol components for telephony applications in Rust.
Fully local AI voice agent (STT->LLM->TTS) for VICIdial over Asterisk AudioSocket. No cloud APIs, no PSTN. faster-whisper + Ollama + Piper, ~0.3s response latency with barge-in.
Dialer-Agnostic AI Voice Agent Platform with Built-in eyeBeam/X-Lite Softphone Hub, Asterisk AudioSocket, VICIdial CTI, and Autonomous Cognitive Brain
To associate your repository with the audiosocket topic, visit your repo's landing page and select "manage topics."