I’ve spent years at the edge where startups meet enterprise adoption — helping new tech break into big companies and watching what actually scales.

One thing I kept seeing: hackathons are fun, but too many weekend demos die because no one knows if they’re building something real or fundable.

So we built *Proving Ground* — a 72-hour “startup sprint” for the AI era:

• Build or ship a real feature in 3 days • Test with real users immediately • Judges score on Pull (market demand), Momentum (shipping speed), and Team Fit (domain know-how) • We track what happens next — who keeps building and who raises

Example: teams from recent sprints have already been invited to 500 Global and are raising seed rounds.

Next sprint is Oct 17–19 in the Bay Area. Open to any builder — no invite hoops, just proof. https://lu.ma/startupsprint

Curious what HN thinks: does this solve what hackathons miss? Would you join if you were validating a new AI idea?