MealMinder

• 

  A photo of our front page!

Inspiration

As university students, we’re all too familiar with overstocked fridges, forgotten leftovers, and roommates letting food go bad. We realized that there wasn’t an easy, automated way to track groceries and their expiration dates. We wanted to create a tool that helps people reduce food waste, save money, and eat fresher by automatically tracking what’s in your fridge.

What it does

MealMinder will combine grocery tracking and AI-powered recipe suggestions. After signing in, users can access their personal fridge page, which will display all stored food items along with the number of days until they expire. Users can input their groceries and we will generate a recipe to use the items in their fridge!

How we built it

Frontend: Built with React and Tailwind, with plans to finish the implementation of Firebase authentication for user accounts and a responsive interface. Users can view their fridge and upload receipts seamlessly.

AI & Automation: We configured a project in Google Cloud Console and obtained a Gemini API key from Google AI Studio, utilizing model version 2.5. By designing a detailed and context-aware prompt, we were able to integrate the Gemini API to dynamically generate recipe suggestions.

Database & Storage: Temporarily stores users in local storage. Firebase Firestore is configured for user inventories and Firebase storage will handle image uploads securely.

Challenges we ran into

The first thing we struggled with was authentication. Implementing Firebase Authentication for the first time required learning new concepts about tokens, user sessions, and secure backend verification that gave us a lot of initial errors. Another new experience was learning OCR and image handling. Parsing the receipts correctly using Google Cloud Vision became difficult due to variations in image quality, formatting, and handwriting. Overall, the backend integration: Building a backend capable of receiving image uploads, running OCR, and sending data back to the frontend was challenging, especially while keeping everything secure.

Accomplishments that we're proud of

Our team worked well together and had a chemistry that deeply increased our enjoyment of the hackathon. Our great collaboration using Git and version control also allowed us to manage the project efficiently during the weekend. We take pride in our clean UI, being able to utilize the Gemini API successfully to create recipes, and the ability to adapt and quickly learn new technical skills.

What we learned

We got practical experience with OCR, text parsing, and cloud APIs. We also learned a lot about backend-to-frontend integration and version control. Additionally, we learned about how important an adaptable, communicative team is -- especially under a time crunch.

What's next for MealMinder

• Implement push notifications to alert users a few days before food expires.
• Expand AI-generated recipes to be more detailed, robust, and customizable.
• Explore categorization and smart suggestions for groceries, so users can plan meals efficiently and reduce food waste even further.
• Further implement OCR functionality

Built With

• css
• gemini-2.5
• html
• javascript
• react
• tailwind