Scan. Decode. Eat Smarter.
NutriScan is an AI-powered food label scanner that cuts through the fine print of packaged food labels.
One snap → instant truth about hidden sugars, additives, allergens, and how processed your food really is.
Food labels are built to confuse:
- Sugar split into 10 different names so it looks “less.”
- Harmless-sounding phrases (apple juice concentrate, natural flavors) hiding junk.
- Numbers & jargon no one understands.
Result: Families, kids, elderly, and health-conscious people are left in the dark.
✨ With NutriScan, food labels finally make sense.
- Snap a Label – Upload or scan any food package
- OCR + AI Analysis – Ingredients decoded & cleaned
- NOVA Health Score – 1 (best) → 4 (ultra-processed)
- Hidden Risk Flags – Sugars, additives, preservatives, allergens tailored to users needs
- Personalized Alerts – Based on allergies & health conditions
- Cart Analysis – See the big picture of your grocery haul
- Individual Analysis: Safety levels (Safe → Caution → Concern → Avoid)
- Hidden Sugar Detection: 50+ sugar variants (maltodextrin, corn syrup, concentrates)
- E-Number Decoder: Translates E102, E621 into plain English
| Group | Description | Health Impact |
|---|---|---|
| 🟢 1 | Unprocessed/Minimal | Healthiest option |
| 🟡 2 | Culinary Ingredients | Use in moderation |
| 🟠 3 | Processed Foods | Limited consumption |
| 🔴 4 | Ultra-Processed | Avoid when possible |
- Allergen Detection: Top 14 allergens + hidden sources
- Additive Analysis: Preservatives, colors, flavor enhancers
- Processing Concerns: Chemical extraction, synthetic compounds
- Health Alerts: Diabetes, hypertension, kidney disease warnings
- FastAPI: High-performance async API server
- Google Gemini: Vision AI for ingredient extraction
- LangChain + FAISS: RAG system for knowledge retrieval
- Vector Embeddings: Semantic search
- llama3: NLP processing pipeline
- React + Vite: Lightning-fast, no framework bloat