Open Endeavour started with a simple itch: everyone talks about what AI will do, and almost nobody shows what one ordinary person can do with it right now.
So that’s the whole premise. Pick something that sounds too big for one person — like training your own embedding model — and do it anyway. In public. With the receipts: the code, the failed runs, the benchmarks, the bills.
Resume Matcher was the first proof. The embedding model is next. The list grows from there.
The rules of the house
01Build in public.Progress, dead-ends, and embarrassing bugs all get documented. Especially the bugs.
02Open everything.Open code, open weights, open notes. If it can’t be shared, it doesn’t belong here.
03No gatekeeping.No “ML PhD required”. If you’re curious, you’re qualified. Come build.
→ sounds fun? the Discord is this way.