built a voice-ai agent using @livekit
handled things like:
- speech to text
- text to speech
- voice activity and turn detection
- tool calling
- fallbacks on models
checkout the demo ๐
so i started building my own harness using providers like gemini, openai and anthropic (not yet for future scope).
and i wanted to explore the voice side of AI so that i can add this dictating feature too.
so right now i build an "ECHO" feature.
- dictating
- audio getting saved (locally)
- TTS (text-to-speech) using openapi and saved audio
working decently good, checkout the demo ๐.