The concept
Culminating project for a full-stack course, solo build. A wellness journal that tracked how the weather lined up with your mood over time. Open it, tap how you’re feeling, write a sentence to future-you. Quick, no friction. The OpenWeather API ran in the background, and everything else lived in a local Django database. ShadCN had just dropped at the time and barely anyone was using it, which is why the whole interface is built around it.
Defending it
The course finished with a live oral exam. The professor sat across from us and asked us to modify features in real time: swap an API, change the weather logic, walk him through the codebase, explain a decision out loud. Intimidating in theory. In practice it was clarifying. You either understood your own code or you didn’t, and you found out fast. A+, and he seemed pleased.
What showed up in the data
Watching patterns emerge in test data was the best part. Seasonal dips that lined up suspiciously well with the calendar. The app wasn’t trying to fix anything, just remind people that what they’re feeling isn’t happening in a vacuum.
If I built it now, post-Avatone, I’d give it more whitespace and far fewer features. Less is generally more. But I was very proud of this one.