Building a Functional Fitness App
An exploration in agentic development.
I’ve used a lot of fitness apps. Some are great if you’re into traditional gym workouts. Others nail running or cycling. Strava is awesome for what it does. Garmin gets the job done if you don’t mind things being a bit traditional. And there are some custom training platforms out there that get close, but they’re purpose-built for a specific program, and that’s about it.
None of them quite fit what I was looking for: something rich in insights, dynamic enough to follow me from the gym to a climbing wall to a trail, and that treats me like an athlete, not just someone logging reps. I wanted something that covered both sides of the equation, the athlete tracking workouts and the trainer programming them.
So I decided to build it myself. And then life happened. I’ve started and stopped this project more times than I’d like to admit. Time has always been the thing that gets in the way.
Why Now?
What changed? AI is reshaping how software gets built, and I wanted to get my hands dirty with it — not just read about it. I’m building this app with Claude Code, Anthropic’s agentic coding tool, and the experience has been a genuine education in what agentic development actually looks like day to day.
On top of that, LLMs open up some genuinely interesting possibilities for a training app; the kind of features that would’ve been impractical to build as a solo developer a couple years ago.
More to come as the project evolves.