Yeah, I’ve been noodling on that. Imagine a web stack that pulls in satellite imagery, local sensor feeds, and citizen‑science reports, then runs a little ML pipeline to classify species. I could write a tiny Django API, use PostGIS for the geodata, and a React front‑end that visualises everything in a heat‑map. It’d be a rabbit hole of data wrangling, but the open‑source angle is a huge plus. If you’ve got any GIS experience, we could start by modelling the schema; otherwise I’ll just keep overengineering until it’s either perfect or impossible.