Inspiration
Starting your fitness journey can feel overwhelming — especially when you're not sure how or where to begin. That’s why we built this workout tracker: to make getting started simple, stress-free, and sustainable. Whether you're feeling unmotivated, stuck, or just tired of starting over, this tool is here to help you take small steps toward big change. No pressure, no judgment — just progress at your own pace. Because every workout, no matter how small, brings you closer to the strongest version of yourself.
What it does
Our project is an AI-powered nutrition assistant and workout planner that helps users understand the food they eat through image analysis. Users can upload a photo of their meal, and our application uses computer vision and generative AI to identify the food items in the image, generate a healthy recipe based on visible ingredients, Estimate macronutrients, store past meals and insights in a connected MongoDB database for tracking.
How we built it
The backend is built with the Flask framework, MediaPipe (for food/object detection), and Gemini AI (for content generation). MongoDB Atlas handles data storage, while the front-end is designed for accessibility across devices—ideal for fitness enthusiasts, beginners, or anyone seeking healthier habits.
Challenges we ran into
Some challenges we ran into were trying to connect the front-end with the back-end securely using FastAPI. Configuring Google Gemini's AI API and prompting the right response. Finding and connecting accurate datasets to our application, and the difficulty of integrating with the MongoDB database.
Accomplishments that we're proud of
Some accomplishments we're proud of are learning how to use and work with Google Gemini's AI and API keys. Successfully linking the front-end with the back-end. Hosting the back-end of our application using Render. Hosting the website on github. Implementing calculator for BMI, fat percentage, and lean mass. Finally, embedding links to motivational videos and quotes, implementing success stories, and creating our own logo.
What we learned
-We learned how to work as a team -Learned how to step up when members became unavailable -We learned how to implement an Ai API key and integrating into our website, calling from the front-end. -Learned how to link front-end with back-end -Learned how to host the back-end using Render -Learned how to make a website using HTML, CSS, JS -Learned how to host a website using Github -Learned back-end coding -Learned how to generate a logo
What's next for FitFusion
FitFusion wants to use camera to identify body fat percentage, create personalized healthy food recipes based on an individuals goal. We want to integrate MongoDb database to save user data and increase calculator accuracy and speed using a wide range of public nutrition datasets as well as user information. Expand on health and wellness features and implement 24 hour online coaching and nutritional recommendations from certified dieticians and nutritionists.
Log in or sign up for Devpost to join the conversation.