Realtime voice AI assistant — streaming STT → LLM → TTS over WebSocket, with barge-in and live translation. Runs fully local (Whisper + Ollama + Piper) or with cloud providers.

Local-first VAD, barge-in, and turn-taking primitives for interruptible voice agents.

llmpipe is a Go library for linear voice/LLM pipelines: processors pass typed frames downstream (or upstream), with async re-entry, barge-in, idle handling, and pluggable STT/LLM/TTS services.

Minimal real-time voice agent on the OpenAI Realtime API — streaming PCM, server-side VAD, function-calling into local RAG, browser + CLI demos with barge-in

Minimal voice AI agent — STT → RAG → LLM streaming → TTS with barge-in. Built to understand how VoxSetu-class systems work.

Realtime voice AI gateway with turn state, interruption handling, provider fallback, degraded state, audit events, runtime evals, Bun, and TypeScript.

TypeScript starter kit for production voice agents: WebRTC, telephony, realtime model APIs, STT/TTS pipelines, barge-in, handoff, evals, and post-call workflows.

STT/LLM/TTS pipeline with barge-in interruption, per-stage latency budgets, and end-to-end voice-to-voice timing. Pipeline overhead p99=1ms; realistic config V2V p50=559ms. 22 tests.