A training app shaped by real use.

I started Train Libre because I wanted to build a real software project and because I wanted a better way to follow my own training. I enjoy going to the gym and want to make progress, but the apps I tried did not show me the metrics I cared about. I did not want to pay for a pile of features I would never use, or turn a training log into another social platform. I wanted to log a workout quickly, see the numbers that matter to me, and export my data whenever I choose.

I want Train Libre to stay a tool. No ads, no gamification, no social feed, and no server I have to run to use my own data. I decide what the app should solve, which features matter, and how the screens and everyday workflows should feel. I work with AI coding agents in natural language, explain what I want to achieve, and think through the options with them. They write most of the code. I rarely write code myself, but the product decisions, priorities, and final call on whether something works are mine.

I study Computer Science at TU Berlin. Data analysis and the structured way software projects are developed have influenced how I think about features and how to decide whether they are finished. I also bring my own experience with training and nutrition. Before I test a change myself, I run automated tests to make sure new work does not break existing features. Then I try it in the simulator with different test data, install it on my own device, and listen to feedback from friends who use the latest beta.