The problem Smart Eats solves

The problem we address a tedious task in personal nutrition management - the difficulty of creating and personalized meal plans that account for individual dietary needs, preferences, and health goals. Many people struggle with the time consuming process of planning balanced meals, finding appropriate recipes, and consistently tracking their nutritional intake. The complexity of nutrition information and the need to personalize it to individual circumstances often leaves people feeling overwhelmed and unable to make changes to their eating habits.

We developed the Smart Eats, to focus on this core problem and solve it in a unique manner which is automated and personalized in a way as per individual needs. It uses Gemini pro 1.5 to generate custom meal plans based on user-specific data, while a multimodal model (1.5 flash) verifies food choices against generated descriptions and the food image uploaded at the specified plan duration. The application also tracks progress, provides macro-nutrient suggestions, and incorporates gamification elements to keep users motivated by having the score card of other users and showcasing it in the leaderboard. By combining these features, it simplifies the nutrition management process, making it easier for users to follow a balanced diet tailored to their unique needs, track their progress, and stay motivated in pursuing their health goals. This comprehensive approach addresses the multifaceted challenges of personal nutrition management in a user friendly and engaging way.

All the data and the generated data points of the meal plan is stored in the firebase real time database to get the result in the structured and proper manner which user can follow easily. When the image is uploaded by the user for the particular meal of that day, the image is stored in the firebase storage and then gets updated in the profile with the image firebase storage URL which is taken into the account for the analysis and score generation.

Challenges we ran into

There were lot much of challenges which we faced during building the application. The problem was the integration of the angular with the flask and then making it intuitively the proper structure into the firebase to store the result of a generated meal plan.

We figured out the best way by iterating and tried testing, which makes it more modular and the scalable now. Then, we faced issue in the integration of the analyze of the meal route as to get the right meal description to be fetched for that proper meal food submission by the user, for the proper analysis but we figured it out by the end by handling it from the frontend and then sending the correct meal details to the route in the backend.

From the various challenges we faced, we got to learn a lot and also overcame it.