FraME

Inspiration

I've been training for two-three years now and haven't seen as much progress as I hoped. With the boom of AI, I saw the potential in utilizing it to help improve not just my physique but for others who couldn't afford a personal trainer.

What it does

FraME uses Gemini 3 to create a personalized workout plan for our users. The user will upload an image of their current photo to our app and we have Gemini process it along with other information to create an extensive workout plan tailored towards the user to bring them closer to their goal. FraME also has progress logging to keep a user consistent and a chat for the user to ask any questions to Gemini for feedback or adjustments to their workout.

How we built it

We built it using React Native and JavaScript for the front-end, a Flask Python API for the back-end, and Supabase for our database. We also used Figma to design the UI and its elements.

Challenges we ran into

Integrating any adjustments from Gemini into the UI was a challenge because we needed to dynamically update the UI based on a Gemini output.

Accomplishments that we're proud of

We are very proud of our seamless integration of Gemini into our UI. For example, when a user wants to adjust it, we dynamically adjust all the UI elements to reflect the change for the user.

What we learned

Both my partner and I have never built a full app from start to finish with a proper front and back-end, so every part of this process was a big learning experience, from setting up our NoSQL Supabase database to connecting Gemini to our front-end.

What's next for FraME

We want to improve the personalization even more, with progress photo tracking and photo overlay so a user can see their progress. We also want to implement nutrition help as well, but we ran out of time for this hackathon.

Built With

• figma
• flask
• gemini
• javascript
• python
• react-native
• supabase