AI · LLM · web · Android
Bilberry
A private health assistant: a meal journal with macros, weight trend, habits, and daily coaching written by a language model.
The challenge
Calorie tracking dies from friction: searching products, weighing, typing. And generic fitness-app advice never refers to what is actually happening in your week.
The solution
A meal can be logged with a photo or a single sentence - the model recognises the contents and computes the macros. The app tracks weight with a trend line, water, habits, and check-ins, and the coach's daily message is written from the user's real data: specific, occasionally sharp, always on point. An optimiser adjusts the calorie target based on actual results.
Scope & technology
React + TypeScript, a Go API, PostgreSQL, LLM integrations (meal recognition, coaching), an Android app, light and dark modes, PL/EN. The app works with private health data, so we don't show its interface publicly.
