Skip to content

lmlassiter/calltranscript

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 

Repository files navigation

Real-Time Call Transcription Demo (Whisper + Twilio)

THe Demo is here Watch the demo

This is a demo project I made a few months ago to showcase the performance of locally run whisper to transcribe live phone conversations on twilio.

The transcriptions are not perfect and adjustments are needed - chunk sizing and segment merging would benefit from further tuning, especially for overlapping or partial speech.

I have only included the py file doing all the heavy lifting. It's a WebSocket server (FastAPI) handling audio decoding, resampling, and transcription. The frontend is just a functional mock interface designed to demonstrate the backends' real-time AI pipeline, not a fully built out front end.

If you’d like to see the frontend work, I’m happy to clean up and share those files — or walk through a separate project with a fully functional UI.

Frontend: Next.js 15 / TypeScript / Tailwind Backend: FastAPI / SSE / Whisper / Librosa / Torch / Twilio Voice API

About

Twilio Whisper demo use case: Agent call script.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages