NutriScan

Inspiration

Reading nutrition labels is often confusing, especially for people with dietary restrictions, allergies, or specific health goals. We were inspired by the idea of empowering individuals to make informed food choices by decoding complex food labels using technology. We wanted to make it easy for anyone—regardless of their background—to understand what they are putting into their bodies.

What it does

NutriScan is an AI-powered web application that lets users upload images of food labels and get instant, intelligent nutritional analysis. It extracts text using OCR, breaks down nutrients, evaluates health impacts, and even supports Indian regional languages. Users can also ask follow-up questions like:

“Is this safe for someone with lactose intolerance?”

“Is this high in sodium?”

“How does this affect heart health?” The app provides contextual answers and citations for all health-related insights, ensuring transparency and credibility.

How we built it

Frontend: Built using React with Tailwind CSS for fast, responsive design. OCR Engine: Integrated Tesseract.js for in-browser text recognition. AI Layer: Used Perplexity AI Multi-language Support: Implemented Indian language detection and parsing using additional OCR configurations.

Challenges we ran into

Handling poor quality images and blurred labels. Ensuring accurate OCR extraction across multiple languages and fonts. Building a flexible AI prompt system that can interpret nutritional data meaningfully. Designing an intuitive UX for both tech-savvy and non-technical users.

Accomplishments that we're proud of

Real-time OCR and AI integration in a browser-only environment. Seamless multi-language label recognition. Building a question-answering system to provide personalized, intelligent health feedback. Generating summarized nutrition reports with trustworthy citations.

What we learned

How to optimize OCR for varied input conditions and languages. The importance of explainability and trust in health-related AI applications. Balancing fast performance with AI processing constraints in the browser. How to create an inclusive product that serves both urban and rural users in India.

What's next for NutriScan

Launching mobile app versions with real-time camera scanning. Expanding support to more languages and food categories. Adding barcode scanning and voice-based label reading. Partnering with nutritionists to refine health recommendations. Introducing user accounts for tracking dietary patterns and generating custom health reports over time.