SureBite - iOS MVP

An iOS app to help people with food allergies and dietary restrictions safely navigate restaurant menus, especially when traveling abroad.

🎯 Features

• Menu Scanning: Use your camera to scan physical menus
• AI Analysis: Powered by Gemini AI to identify allergens and dietary conflicts
• Risk Assessment: Dishes categorized as Safe, Caution, or Avoid
• Chef Card: Translated allergy warnings to show waiters
• Order Builder: Select safe dishes and create an order list
• Profile Management: Customize allergens, diets, language, and location

📱 Screenshots

See the /UX folder for design mockups.

🚀 Getting Started

Prerequisites

• macOS with Xcode 15+ installed
• iOS 16+ device or simulator
• Google Gemini API key (Get one here)

Setup Instructions

1. Create a new Xcode project:

   • Open Xcode
   • File → New → Project
   • Choose "App" under iOS
   • Product Name: SureBite
   • Interface: SwiftUI
   • Language: Swift
   • Save in /Users/yiransi/Documents/SureBite

2. Add the source files:

   • Drag the entire SureBite/ folder into your Xcode project
   • Make sure "Copy items if needed" is checked
   • File structure should be:

     SureBite/
     ├── Models/
     │   ├── User.swift
     │   └── MenuItem.swift
     ├── Services/
     │   ├── Secrets.swift
     │   ├── OCRService.swift
     │   └── GeminiService.swift
     ├── ViewModels/
     │   ├── ProfileViewModel.swift
     │   └── ScannerViewModel.swift
     ├── Views/
     │   ├── OnboardingView.swift
     │   ├── ProfileView.swift
     │   ├── MenuScannerView.swift
     │   ├── MenuResultsView.swift
     │   ├── DishDetailView.swift
     │   ├── ChefCardView.swift
     │   └── SettingsView.swift
     ├── SureBiteApp.swift
     └── MainTabView.swift
     

3. Add your API key:

   • Open SureBite/Services/Secrets.swift
   • Replace YOUR_GEMINI_API_KEY_HERE with your actual Gemini API key

4. Configure camera permissions:

   • In Xcode, select your project in the navigator
   • Select the target → Info tab
   • Add these keys to the Custom iOS Target Properties:
     • Privacy - Camera Usage Description: "We need camera access to scan menus"
     • Privacy - Photo Library Usage Description: "We need photo access to analyze menu images"

5. Build and run:

   • Select your target device/simulator
   • Press Cmd+R to build and run

🧪 Testing

Test the Flow

1. Launch app → Complete onboarding (select allergies like "Peanuts", "Shellfish")
2. Go to Profile → Verify allergens appear as blue pills
3. Tap "Scan Menu" → Take a photo of a menu (or use sample image)
4. Wait for analysis → View results with risk colors
5. Tap a dish → See detailed risk explanation
6. Add dishes to order → Tap "Show to Waiter"
7. View Chef Card with allergy warning

Sample Menus for Testing

You can create a simple test menu document with items like:

• Grilled Salmon with Lemon
• Pad Thai with Peanut Sauce
• Caesar Salad with Croutons
• Shrimp Tempura
• Vegan Buddha Bowl

📊 Architecture

Built using MVVM pattern:

• Models: Data structures (User, DietaryProfile, MenuItem)
• Services: OCR (Vision framework) and Gemini API client
• ViewModels: ProfileViewModel, ScannerViewModel
• Views: SwiftUI views matching UX mocks

🔐 Privacy

• All user data stored locally on device (UserDefaults)
• Only extracted text sent to Gemini API, not images
• No account/authentication required for MVP

🗺️ Roadmap

See PRD.md for full feature roadmap including:

• Phase 2: Restaurant database, crowdsourced validations
• Phase 3: AR overlay, barcode scanning, social sharing

📝 Notes

• The Secrets.swift file is excluded from git (see .gitignore)
• Gemini API is free for moderate usage (you may need to monitor your quota)
• For production, consider adding a backend proxy to hide API keys

🐛 Troubleshooting

OCR not working?

• Ensure good lighting and clear menu text
• Make sure camera permissions are granted

API errors?

• Check your API key in Secrets.swift
• Verify internet connection
• Check Gemini API quota limits

Build errors?

• Clean build folder (Cmd+Shift+K)
• Restart Xcode
• Verify all files are added to target

📄 License

This is an MVP/prototype project.

---

Built with ❤️ using SwiftUI and Gemini AI