A live deployment can be found at: https://call-analysis-official.vercel.app/
- AssemblyAI for core transcriptions and & audio intelligence
- OpenAI for embeddings and LLM completions
- Supabase for hosting & edge functions
- In ./backend, you'll need to configure your deno settings and familiarize yourself with the supabase CLI to use functions. A good resource on this can be found here
- Get your Supabase DB URL and Anon Key and add it to your functions.
process-videohandles all uploads, transcriptions, and conversion of AssemblyAI's automatically generated chapters into embeddings which are stored & associated with a user's ID. To use these functions, you'll need both an OpenAI key and AssemblyAI key - In ./frontend, install deps with
npm install - Add your Supabase URL and Anon Key as .env variables
- Run
npm run startto start project.