Skip to content

michaelmccrae/deparley

Repository files navigation

Deparley is online instruction for negotiation using practial scenarios. Emphasis is on learning by doing. Deparley is frictionless service to perform various sales calls and negotiation scenarios. Artificial intelligence is seeing growth in services that use human voice for interaction. Deparley is harnessing the best of those services.

Service is forked from Vercel's Swift – AI Voice Assistant

AI and APIs

Service relies on a number of AI APIs:

  • Swift for fast AI voice assistance.
  • Groq is used for fast inference of OpenAI Whisper (for transcription) and Meta Llama 3 (for generating the text response).
  • Cartesia's Sonic voice model is used for fast speech synthesis, which is streamed to the frontend.
  • VAD is used to detect when the user is talking, and run callbacks on speech segments.
  • Deepgram is used for modeling pauses and filler words in transcription.

Tech stack

The app is a Next.js project written in TypeScript and deployed to Vercel. The site also uses Supabase for the Postgresql database.

How it all works

Select a scenario

Select a scenario

Read scenario and press start when ready to practise

Read scenario and press start when ready to practise

After practise, you are measured on asking qualifying questions, use of filler words and pauses

After practise, you are measured on asking qualifying questions, use of filler words and pauses

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors