activesince May 24, 2026

Starting Edentic V2

so i am making one of first wins in a hackathon - Edentic which was based on videodb, into a local ai first editor powered by cli agents or byok providers

5 updates

  1. Going with Rust, Tauri for the desktop app they are much faster and efficient than electron plus rust and the python engine for the editor will be good stable combination

  2. i need to find a way to get the audio transcription, and scene analysis done but without videodb this time

  3. audio is mostly fine whisper for transcribing and librosa for sentiment analysis and cinematic application, did the same thing but on the cloud in v1 so thats good

  4. the problem is with the video indexing, I am not using videodb anymore, so scene analysis is going to be diffcult

  5. i could go with the old gemini method or sending frames to the ai but thats gonna be costly, eat up credits and be inefficient for longer videos say like a 6 hour dump of footage