Frank
An AI fitness app, built and launched in about 1.5 days
A real product, from idea to live in about a day and a half.
1.5 days
idea to live
4
signups in 24 hours
1
AI function, to control cost
The situation
I wanted to show how fast a real product can go live without cutting corners on cost or monitoring.
What I did
- Used AI for one job only, photo analysis, and kept everything else rule-based, so costs stay low and results predictable.
- Built login, email flows, analytics and error tracking from day one.
- Launched with one LinkedIn post.
Why it worked
Clear scope, and AI only where it earns its place.