[Technical Documentation] Qook Architecture & Implementation

1. Tech Stack

• Language: Kotlin
• Architecture: Google App Architecture (MVVM + Repository)
• Dependency Injection: Koin
• Local Database: Room
• Asynchronous Processing: Coroutines & Flow
• AI Model: Google Gemini 2.5 flash
• Subscription Management: RevenueCat SDK
• Backend/Auth: Firebase

---

2. System Architecture

We followed Google’s recommended app architecture to ensure scalability and maintainability.

1. UI Layer: Uses ViewModel and StateFlow to handle UI states and react to data changes.
2. Domain Layer: Contains business logic and use cases for recipe processing.
3. Data Layer: * Repository Pattern: Acts as a single source of truth, mediating between Network (Gemini) and Local (Room) data sources.
   • Room DB: Stores saved recipes and shopping lists for offline access.

---

3. RevenueCat Implementation

We integrated RevenueCat to simplify the complex in-app purchase lifecycle.

• Entitlements: We defined a Qook Pro entitlement that unlocks "Unlimited AI Extractions."

• Dynamic Paywalls: Used RevenueCat to display optimized paywalls, checking user status before allowing Gemini-powered extractions.