NutriScan AI

AI-powered early detection and affordable nutrition planning for malnutrition in underserved communities.

Comment

🌍 Inspiration

Over 45 million children under the age of 5 suffer from wasting (severe malnutrition), and nearly 149 million are stunted globally (WHO, UNICEF). In countries like India, early detection is often missed due to lack of access to healthcare, awareness, and diagnostic tools.

Malnutrition is not just a food problem — it’s a detection and accessibility problem.

If we can identify risk early and suggest affordable, locally available nutrition, we can prevent irreversible damage and save lives.

This inspired us to build NutriScan AI.

💡 What it does

NutriScan AI is an AI-powered platform that enables:

  • Image-based malnutrition detection (via computer vision or mock AI)
  • Symptom-based risk prediction for low-tech environments
  • Affordable meal recommendations using locally available foods
  • Government scheme integration (Mid-Day Meal Scheme, ICDS)
  • Offline-first usability for rural and underserved regions

It bridges the gap between detection → decision → action.

⚙️ How we built it

  • Frontend: React + Tailwind CSS (mobile-first UI)
  • Backend: Node.js / FastAPI
  • AI Layer:
    • Image classification (mock or lightweight model)
    • Rule-based + ML logic for symptom analysis
  • Data Logic:
    • Nutritional mapping of low-cost foods
    • Cost estimation engine (₹ per day diet plans)
  • APIs (optional): OpenAI API

We focused on speed, clarity, and scalability, ensuring the solution works even with limited internet access.

🧠 What we learned

  • Building for underserved communities requires constraint-first thinking
  • AI doesn’t need to be complex to be impactful
  • UX matters deeply in low-literacy environments
  • Scalability comes from simplicity and accessibility

🚧 Challenges we ran into

  • Lack of real malnutrition datasets → solved using mock AI logic
  • Designing for low bandwidth environments
  • Balancing accuracy vs speed within hackathon time
  • Making recommendations affordable yet nutritious

🚀 Impact & Scalability

  • Can be deployed in rural India, Sub-Saharan Africa, and Southeast Asia
  • Works on low-cost smartphones
  • Can integrate with NGOs, government programs, and health workers

Even basic early detection can reduce severe malnutrition cases by enabling timely intervention.

🔮 What’s next

  • Integrate real computer vision models
  • Add multilingual support (Hindi + regional languages)
  • Partner with NGOs and governments
  • Build a community health monitoring dashboard

❤️ Why this matters

Malnutrition is preventable — but only if we act early.

NutriScan AI empowers communities to detect, act, and nourish.

Built With

Share this project:

Updates

Submission history