Inspiration
Walking through grocery stores, we realized how hard it is to make informed decisions about health, sustainability, and dietary requirements. We wanted to make conscious consumption as simple as scanning a barcode.
What it does
ScarletScanner instantly analyzes products via barcode scanning:
- Nutri-Score & Eco-Score with AI explanations
- 8 sustainability metrics (greenhouse gas, water, labor, animal welfare, etc.) scored 0-100
- Auto-detects Vegan/Halal/Kosher from ingredients
- Suggests healthier alternatives
- Finds local farmers markets by ZIP code
How we built it
- Backend: Flask + Groq AI (Llama 3.3 70B)
- Data: Open Food Facts API, OpenStreetMap
- Frontend: JavaScript + Quagga.js for barcode scanning
- Methodology: HowGood's 8 equally-weighted sustainability metrics
Challenges we ran into
- Groq API JSON parsing errors (solved with
response_formatparameter) - Handling decimal scores vs integer ranges for color coding
- Building accurate ingredient analysis for dietary certifications
- Managing multiple API rate limits
Accomplishments that we're proud of
- Zero-cost local market finder using only free APIs
- Real-time camera barcode scanning with validation
- Intelligent dietary detection from actual ingredients
- Beautiful color-coded sustainability visualizations
What we learned
- Prompting LLMs for consistent JSON output
- HowGood sustainability methodology
- Geospatial queries with Overpass API
- Importance of visual design for complex data
What's next for ScarletScanner
- Native mobile apps with offline mode
- Shopping cart sustainability tracking
- Community reviews and crowdsourced data
- Personal carbon footprint dashboard
- Allergen detection and retailer partnerships
Log in or sign up for Devpost to join the conversation.