AI helped me try a new workout app
I have been training in the gym for a few years now, mostly doing strength training. A nice side adventure that comes with it is trying out new apps to track my workout metrics. While the real purpose is doing the actual training, I cannot deny the joy of trying new apps and tech, and admit that it motivates me to do even more.
For the past two and a half years, I have been using the Hevy app. It is a great app with a great UI and Apple ecosystem support (Live Activities, Health app, and Apple Watch integration), and it is frequently updated, which is always a good sign. To top all of that, most of the app features are free to use (although I am happy paying for Hevy Pro, as I love paying for the things I use).
That being said, I can’t resist the urge to try new apps from time to time. At the beginning of January, MacroFactor Workouts came out and I really wanted to try it. Since I subscribe to their nutrition app, I got the first year free as part of a bundle promotion for the new release. I spent the past month trying it out, but unfortunately it wasn’t enough for me and I missed some Hevy features. So for now I’m back to Hevy. The problem was that all my logged workouts were now stuck in MacroFactor, and I wanted that history in Hevy too.
There wasn’t an existing tool that did this, so I tried to make one for myself with some help from Claude Code and Codex. This is how I ended up with: Workout Data Sync — a website I made to convert data from MacroFactor to Hevy. I originally built this mainly for myself as a simple Python script, but then I thought, why not make it publicly available? So I pushed a bit further, made everything open source, and created a website that you can also run locally if you want.
The usage is quite simple: you export your history from MacroFactor, get a CSV file, upload it to the website along with your Hevy API key (Pro required), resolve any exercise name differences, and import. One annoying part was matching exercises that have different names between the two apps. The tool handles the common ones (based on what I encountered) automatically based on naming similarities and asks you to choose for anything it doesn’t recognize.
Needless to say, I probably wouldn’t have done it without AI — it would have been too much effort for the small inconvenience of losing one month of workout history. But not only did I end up doing it, I even had fun in the process!