Show HN: I indexed 37h of my videos using an RTX 4090 and local ML models in 24h

Category: ai-ml

Tags: video-indexing, machine-learning, self-hosted

Score: 4.3/10 (Innovation: 4, Technical: 5, Documentation: 3, Utility: 4)

This project uses local ML models on an RTX 4090 to index long-form video content (podcasts, tutorials, screen recordings) by analyzing frames for faces and text, all run via Docker. Its interesting for demonstrating practical self-hosted video indexing with consumer hardware, though the documentation is minimal and the approach is incremental.

Target audience: data engineers

View on Hacker News