ReCovr

Inspiration

We were inspired by the miraculous recoveries of NBA players(and also the downfall of Tyrese Haliburton). We saw their world-class rehabilitation and thought that everyone deserves their level of care and consistency, so we made a web app to help with logging and give baseline advice.

What it does

Our project aims to make rehabilitation easier through giving tips, encouragement and exercises. Additionally, it also facilitates better communication with doctors as you can easily show your data/progress through the concise summary or graphs.

How we built it

We drafted our project using Figma, choosing light colours that represent the healing journey. We chose to make a Web App for wider compatibility (both desktop and mobile). We built our project through next.js, tailwind, ShadCN for the components, and Google Gen AI for the Gemini Model that provides suggestions. After finishing our app, we hosted it using Vercel.

Challenges we ran into

On Vercel, we were unable to connect our Gemini AI key, which halted our progress. Since most of us had little React experience, we also had to learn along the way, which led to many basic mistakes. Another problem was time management, as we spent so much time on designing our site instead of programming it and its functions!!!

Accomplishments that we're proud of

We're proud that we were able to finish our project with our novice-level hackathon/programming experience. We were also content with our site design and how we implemented our Figma design.

What we learned

Two members came into the project with NO React experience; afterwards, they learned the React fundamentals, Tailwind and how to control vibe-coding. We also learned how to use Vercel to ship our project, and how to deal with Git merge conflicts!!!!

What's next for ReCovr

Since our suggestions are general (as the AI doesn't have substantial info), we could train models to identify bruises, soreness and use OpenCV to measure flexibility from video footage; these extra data points would allow the AI to provide more accurate suggestions.

Additionally, we could improve our login to be compatible with Vercel and work to develop a dedicated mobile app(working cross-platform).