Inspiration
The project was inspired by personal experiences and anecdotes where we noticed shortcomings in current fitness knowledge and existing applications for tracking workout progress.
What it does
gAI.nz is an AI-powered workout planner that turns your goals, weekly progress photos, and real workout stats into a personalized training + nutrition plan that actually adapts as you do. Instead of following a static “bro split” or rigid science template, it learns from what you completed, how you felt, and where you’re plateauing—then updates your regimen and nudges you to switch things up when needed. The result is a heavily customizable, themed plan with real accountability that keeps weight loss or muscle gain on track.
How we built it
We used React Native and Tailwind for the frontend. For the backend, we used a Postgres database, Flask API, and the Claude API.
Challenges we ran into
We ran into issues using the Claude API since we only had access to the free tier. This meant we were rate limited to 5 requests per minute, so we had to make certain implementation decisions
We had some API calls grouped together in functions which resulted in long wait times that were blocking changes from rendering on the frontend. To solve this we split our LLM calls into separate functions when possible and loaded changes asynchronously.
Built With
- claude
- docker
- flask
- postgresql
- python
- react
- react-native
- tailwind
Log in or sign up for Devpost to join the conversation.