AI Ingredient Scanner
A multi-agent AI system for analyzing food and cosmetic ingredient safety. Features a Streamlit web interface, React Native mobile app, and RESTful API powered by Google Gemini 2.0 Flash and LangGraph orchestration.
What It Does
The AI Ingredient Safety Analyzer helps users understand product ingredient safety by:
- Parsing ingredient lists from any food or cosmetic product
- Researching each ingredient via vector database and Google Search
- Generating personalized safety reports based on user allergies and skin type
- Providing actionable recommendations (SAFE / CAUTION / AVOID)
Key Features
Core Capabilities
- ✅ Personalized safety analysis based on allergies and skin type
- ✅ Beginner/Expert explanation modes
- ✅ Allergen matching with prominent AVOID warnings
- ✅ Quality validation with automatic retry loop
- ✅ Session persistence across page refreshes
Mobile App (Phase 2)
- 📱 Native camera integration for label scanning
- 🌍 Multi-language OCR with auto-translation (9+ languages)
- 🌓 Dark/Light theme toggle
- 📋 Expandable ingredient cards with detailed safety metrics
Supported OCR Languages
English, French, Spanish, German, Italian, Korean, Japanese, Chinese, Portuguese
Architecture Preview
Screenshots
| Web Interface | Mobile App |
|---|---|
| Streamlit dashboard with safety reports | React Native app with camera scanning |
Quick Links
Repository
📦 GitHub: udaytamma/AiIngredientScanner
Version History
| Version | Description |
|---|---|
| v2.0.0 | Mobile app, REST API, multi-language OCR |
| v1.0.0 | Initial release with Streamlit web interface |