RowdyHacks XI 1st Place 🏆

🚀 Setup & Installation

Prerequisites

• Node.js (v16+), Python (v3.8+), and OpenRouter API key

Backend Setup (Run from backend/ directory)

cd backend
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate
pip install -r requirements.txt
cp .env.example .env
# Add your OpenRouter API key to .env: OPENROUTER_API_KEY=sk-or-v1-xxxxx
uvicorn main:app --reload --host 0.0.0.0 --port 8003

Frontend Setup (Run from frontend/allergen-app/ directory)

cd frontend/allergen-app
npm install
cp .env.example .env
# Add your machine's IP to .env: EXPO_PUBLIC_API_BASE_URL=http://YOUR_IP:8003
npx expo start

Find your IP:

• macOS/Linux: ifconfig | grep "inet "
• Windows: ipconfig

Testing:

• Press i for iOS Simulator (macOS)
• Press a for Android Emulator
• Scan QR with Expo Go app (ensure same WiFi network)

Verify Backend: Visit http://localhost:8003/docs for API documentation

💡 Inspiration

Food allergies can turn a simple grocery trip into a stressful guessing game. Our team wanted to make that process easier, safer, and faster — especially for people who constantly need to double-check ingredients or risk severe reactions.

We created an app allowing you to scan any product, instantly know if it’s safe, and get smart alternatives powered by the gold-standard of food databases, supplemented by ChatGPT. That’s ALERG'Z — a personalized food companion built to bring peace of mind to those with allergies 👍

---

⚙️ What it does

ALERG'Z is a mobile app that helps users identify allergy-safe foods with just a barcode scan.

Here’s how it works:

Step	Description
1. User Profile Setup	Users input their allergens when they first open the app, user data is stored with an SQLite database
2. Product Scanning	When a barcode is scanned, ALERG'Z instantly fetches data from OpenFoodFacts to identify the product and its ingredients.
3. Allergen Detection	The backend cross-checks ingredients against the user’s allergy profile.
4. AI-Driven Alternatives	If a product isn’t safe, ALERG'Z uses ChatGPT (via OpenRouter) to suggest similar, allergy-friendly alternatives tailored to the user’s region.
5. Visual Feedback	Each recommendation includes live-fetched product images for a smooth, engaging experience.

All of this happens in seconds, giving users clear, visual answers and personalized options — no ingredient lists or guesswork required.

---

🛠️ How we built it

• Frontend: A sleek, modern React Native interface designed for both speed and accessibility. We focused on a clean user experience where every tap feels instant and intuitive.
• Backend: Built with FastAPI in Python, enabling rapid, concurrent API calls to OpenFoodFacts and GPT’s generative model for alternative recommendations.
• APIs & Data:
  • OpenFoodFacts API for product details and UPC lookups
  • OpenAI API for intelligent food alternative generation
  • Pixabay / OpenFoodFacts images for visual product data
• Architecture:
  • Asynchronous background threads to fetch product images dynamically
  • JSON-structured responses to sync data seamlessly with the React Native client
  • User authentication and profile management using SQLAlchemy + FastAPI ORM

This setup allowed us to achieve real-time scanning and recommendations — even on limited bandwidth.

---

🚧 Challenges we ran into

• Image Updating: Allowing dynamic loading of page elements to ensure a fast, accessible user experience.
• API Rate Limits: Balancing performance while staying within OpenFoodFacts and ChatGPT rate constraints required optimization and caching logic.
• Prompt Engineering: Tuning the AI prompt for consistent, list-formatted responses from ChatGPT took a lot of iteration.
• Cross-Platform Sync: Ensuring React Native and FastAPI communicated perfectly across threads and sockets pushed us to fine-tune CORS and websocket behavior.

---

🏆 Accomplishments that we’re proud of

• Built a fully functional allergy-detection and recommendation system from scratch in just a few days.
• Achieved real-time barcode scanning and analysis with a fast, responsive UI.
• Successfully integrated multiple APIs and AI models into one seamless experience.
• Designed a clean, user-first mobile interface that feels both modern and trustworthy.

Most importantly, we created something that could genuinely make life easier for people managing allergies every day.

---

📚 What we learned

• How to efficiently orchestrate multi-API workflows in FastAPI using background tasks and concurrent threading.
• The power of prompt design — small phrasing changes can dramatically impact AI output.
• Strategies for frontend-backend synchronization in real-time, user-centric applications.
• How valuable user empathy is when building healthcare-adjacent products — the human side matters just as much as the technical side.

---

🚀 What’s next for ALERG'Z

• Ingredient-level risk scoring to show how “safe” a product is for a specific allergy.
• Expanded database support for region-specific food APIs (USDA, Tesco, Carrefour).
• Camera-based label scanning for offline use when barcodes aren’t available.
• Personalized dashboards to track safe foods, allergen exposure, and nutrition trends.
• Social recommendations, allowing users to share safe product discoveries.

Our vision is to make ALERG'Z the go-to app for allergy-conscious consumers — empowering users to eat confidently, anywhere.

📁 Repo Structure

ALERGZ
├── README.md
├── LICENSE
├── .gitignore
├── backend
│   ├── main.py
│   ├── requirements.txt
│   ├── allergen.db
│   └── app
│       ├── api
│       │   ├── scan.py
│       │   ├── recommend.py
│       │   └── __init__.py
│       ├── database
│       │   ├── models.py
│       │   ├── db.py
│       │   └── __init__.py
│       ├── schemas
│       │   └── product.py
│       └── __init__.py
├── frontend
│   └── allergen-app
│       ├── package.json
│       ├── tsconfig.json
│       ├── babel.config.js
│       ├── eslint.config.js
│       ├── .env
│       ├── app
│       │   ├── _layout.tsx
│       │   ├── (tabs)
│       │   │   ├── screen1.tsx
│       │   │   ├── screen2.tsx
│       │   │   ├── screen3_camera.tsx
│       │   │   └── explore.tsx
│       │   └── components
│       │       ├── GlassSearchBar.tsx
│       │       ├── SelectedItemsArea.tsx
│       │       └── ui
│       │           └── icon-symbol.tsx
│       ├── assets
│       │   └── (images, icons)
│       └── scripts
│           └── reset-project.js
├── docs
│   └── (design / hackathon notes)
└── misc