I built and released an Apple Watch tennis scoring app with AI as my development partner
Hi everyone,
I wanted to share a small but real project I’ve been building with AI assistance: an Apple Watch app for tracking tennis matches directly from the wrist.
The app is called Tennis Score Wizard. It tracks tennis scoring rules: points, games, sets, tie-breaks, singles/doubles, server position, undo, match history, and Apple Health workout recording with heart rate support.
What made this project interesting for me is that I used AI not just for random code snippets, but almost like a development partner:
- planning the app structure
- designing the watchOS UI for small screens
- debugging scoring edge cases
- writing and refactoring Swift code
- improving App Store metadata
- thinking through future features like match statistics, iCloud backup, and monetization
The hardest part was not “writing code with AI”. The hardest part was keeping the product simple. On Apple Watch, every extra tap matters. For example, we discussed adding detailed tennis stats like aces, double faults, winners, unforced errors, first serve percentage, etc., but decided not to do it because it would require extra input after every point. The goal is still: tap who won the point and continue playing.
The next feature I’m planning is automatic match statistics in History, based only on the normal scoring taps:
- total points played
- points won %
- service games won
- return games won
- service points won
- return points won
- tie-breaks won/lost
- longest game
- average points per game
No extra taps, no complex charting workflow.
This was my first real App Store release, and it was interesting to see how AI helped not only with implementation, but also with product decisions, UX trade-offs, release planning, and even App Store review preparation.
App Store link:
https://apps.apple.com/app/tennis-score-wizard/id6782709004
I’d be curious to hear feedback from other developers:
How do you keep AI-assisted development from turning into feature creep?
And do you think small utility apps like this should stay paid upfront, or move to free + one-time Pro unlock?