Hey all launching Burgndy.ai today. Quick context: I come from ops/QA, not engineering, and about a month ago I got frustrated with every "AI app builder" I tried either giving me a web page pretending to be an app, or code that looked plausible and then fell apart the second I tried to actually run it on a device.
So I built one that doesn't let that happen it compiles a real Flutter app, tests it on an actual Android emulator/iOS Simulator, and can push straight to the Play Store / App Store. There's also a chat-driven "Studio" mode now where you just describe what you want and watch it get built in a live sandbox.
Two things I'm genuinely unsure about and would love brutally honest feedback on:
1. Does "it actually compiles and runs on a real device" land as a real differentiator, or does everyone assume that's already table stakes?
Quick update, 8 hours in: real, useful responses already coming in, thank you to everyone who's tried it or left thoughts. Still especially curious about question 2 above — if you've been burned by a vibe-coded app before, what actually broke? That's the exact thing I'm trying to get right, not just the demo version of it.
If you've tried Burgndy.ai today, an honest upvote or a quick comment (even "this doesn't solve anything for me because X") genuinely helps more than a polite pass — trying to actually learn something from today, not just get numbers up.
Another update: a tester hit a real bug within hours of launch (iOS build failing on an expired credential) and reported it directly. Fixed it, verified it end to end with them, and while I was in there found and closed a bigger gap too, the Android publish flow was pointing at a step that didn't actually work yet. Both are live now.
Genuinely appreciate everyone who's tried this today and told me what broke instead of just bouncing. That's more useful to me right now than any upvote.