GreenScore

GreenScore helps you scan or search food products to instantly see how sustainable and ethical they are, with a simple AI score based on environmental impact, labour, and transparency.

Comment

🌱 Inspiration

We were inspired by the growing movement toward ethical consumerism and sustainable living, but also frustrated by how difficult it is to make truly informed choices. Greenwashed packaging, vague eco-labels, and hidden supply chain issues often leave well-intentioned shoppers in the dark. GreenScore was created to bring clarity and accountability to the food industry, giving consumers the power to align their purchases with their values in a simple and confident way.

🤖 What it does

GreenScore helps users make ethical and sustainable food choices by scoring products out of 100 based on environmental impact, labor practices, ingredient sourcing, and overall transparency.
We also take into account additives, preservatives, and other health-related factors to give a more well-rounded evaluation of the product.

Users can simply scan a barcode or search a product name, and in seconds, our AI analyzes verified data and returns:

  • A GreenScore (0–100)
  • A concise summary of sustainability, ethics, and health impact
  • A list of pros, cons, and verified facts
  • Suggested similar products with better scores

To guide users even further, we include an AI-powered chatbot assistant that helps answer questions, recommend better options, and support smarter, more values-aligned choices while shopping.

🛠️ How we built it

We built GreenScore with a modular, full-stack architecture to ensure fast performance and scalability:

  • Frontend: React + Tailwind CSS for a sleek and responsive UI
  • Backend: FastAPI with structured prompt engineering for Gemini AI integration
  • Database: MongoDB Atlas to store user profiles and search history
  • Authentication: Firebase Auth to handle secure user login and access

Our scoring logic is powered by prompt-engineered responses from Gemini AI, which analyzes product attributes and maps them to a transparent scoring rubric.

🧱 Challenges we ran into

  • Sourcing reliable, structured data for sustainability and ethics that could be evaluated consistently
  • Designing AI prompts that returned factual, parsable, and bias-free summaries
  • Ensuring fast, accurate barcode scanning in browser across devices and platforms
  • Building a clean, intuitive experience that works under hackathon time constraints

🏆 Accomplishments we're proud of

  • A functional product with real-world use cases
  • Successfully integrated AI for complex scoring with consistent structure
  • Created a meaningful user experience that aligns tech with impact

📚 What we learned

  • How to harness LLMs in a controlled, reliable way for real user applications
  • The importance of frictionless design in driving adoption of sustainability tools
  • How to work efficiently as a team to divide, conquer, and polish a full-stack AI project in under 48 hours

🚀 What’s next for GreenScore

  • Expanding the product database with verified and crowdsourced sustainability data
  • Integrating *MCP * to keep scores and suggestions in sync with real-time product updates
  • Integrating a browser extension for instant feedback while shopping online
Share this project:

Updates