This demo application uses the `incredibly-fast-whisper` (Replicate) to transcribe audio, LLaMA 3 (Groq) to summarize, Supabase database to store data and Cloudflare R2 to upload audio files.
Audio is uploaded into Cloudflare R2, then transcribed and the data is stored in your Supabase PostgRES database.
The main frontend logic is found in the app/(apps)/audio folder.
The main configuration file can be found in app/(apps)/audio/toolConfig.ts file.
The API transcription endpoint and logic can be found in app/api/(apps)/audio/transcribe/route.ts.
The API to summarize the transcript be found in app/api/(apps)/audio/summarize/route.ts.
The API to upload audio can be found in app/api/(apps)/audio/upload/route.ts.
The API to delete audio & Supabase entries can be found in app/api/(apps)/audio/delete/route.ts.