NutriScan AI is a web application built in 36 hours at HackWestern 12 that transforms a simple selfie into personalized nutrition guidance. Using multimodal AI, NutriScan detects subtle physical cues linked to nutritional imbalances and generates a tailored, budget-friendly meal plan to help users improve their health.
Over the weekend, our team won:
🏆 Sun Life's Best Health Hack 🏆 MLH Best Use of DigitalOcean Gradient AI
Features
- Visual Health Screening: Upload a selfie or hand photo to detect visible nutritional markers.
- AI-Generated Meal Plans: Personalized diet recommendations tailored to student budgets and needs.
- Real-Time Analysis: Powered by Google Gemini 2.5 Flash and DigitalOcean Gradient AI.
- Privacy-Focused: Images are processed per session only and are never stored.
- Clean, Accessible UI: Built with modern frontend tooling.
The Problem Students often sacrifice health and nutrition for deadlines. Early signs of poor nutrition show up physically -> pale skin, brittle nails, fatigue cues, but most of us don’t know how to interpret them. This leads to burnout, poor overall energy, and long-term health issues.
Our Solution NutriScan AI turns a basic photo into actionable insight. Using Google Gemini's multimodal vision API, the system identifies potential nutritional issues and instantly generates a practical, affordable meal plan designed to help users course-correct quickly.
Why It Matters Many students don’t know where to begin when it comes to improving their nutrition. NutriScan AI helps to lower that entry barrier by:
- Making health cues easily understandable
- Offering realistic, personalized advice
- Encouraging small, actionable habits
- Empowering users before issues escalate
- This is a step toward AI-driven preventative health that anyone can access.
Our Tech Stack Frontend
- React
- TypeScript
- Tailwind CSS
AI + Backend
- Google Gemini 2.5 Flash (Multimodal Vision)
- DigitalOcean Gradient AI
- Node.js / Express
- Vite
Deployment: DigitalOcean
Test Out Live At https://nutriscans.tech