inspired by a video I watched about a footage crawler

Hi! I'm new here and wanted to ask a genuine question — not sure if this is already solved somehow or if it's something I'd need to build from scratch.

I'm exploring an idea for Dad Labs (a community for dads learning to use AI) — inspired by a video I watched about a footage crawler: reducing hours of video into searchable text descriptions using extracted frames plus a local vision model, all organized through folders/markdown.

The application I have in mind is different from the one in the video (which was a 9TB professional archive) — I'm thinking something much smaller: helping families organize the years of video they've accumulated (weddings, first steps, etc.) that they never have time to sort through.

I don't know if this is something the community already has as a template, or if it's more of an end-to-end project I'd need to research and build myself. I'd rather ask than assume — is there something similar already documented here, or any reference you'd recommend to start understanding the actual technical structure (folder organization, the markdown file format, etc.)?

Any guidance is welcome, and if this is an obvious question for anyone who's been here longer, apologies — I'm just getting started.

5 comments

inspired by a video I watched about a footage crawler