Many gym goers struggle to maintain consistent training progression, without a clear way to track and review their workouts. Current applications are either too basic or too complex, leaving a gap for an accessible, data driven training application. The project is a cross platform mobile gym tracking application, built with React Native and Expo. This project uses offline-first with SQLite so it’s in a gym environment without internet connection. A rulebased recommendation engine analyses the user's training history. It applies progressive overloading principles to suggest weight adjustments and which muscle groups to train. There is also an opt-in AI coaching mode that connects to GPT-4o through the OpenRouter API.