📸 MacroSnap

MacroSnap is a modern, AI-powered nutrition tracking application designed to simplify healthy eating. By leveraging Google's Gemini AI, the app allows users to snap a photo of their meal and instantly receive a detailed nutritional breakdown, including calories and macronutrients.

---

🚀 Key Features

• 🤖 AI Meal Analysis: Integration with Google's Gemini 2.5 Flash model to identify dishes and estimate calories, protein, carbs, and fats from a single image.
• 📊 Weekly Dashboard: A comprehensive summary of weekly intake, helping users stay on track with their nutritional goals.
• 📂 Meal History: A dedicated log of all scanned meals with sorting options (Date, Alphabetical) and local persistence.
• 🔐 Secure Authentication: Implements Firebase Authentication coupled with the modern Android Credential Manager for a seamless and secure sign-in experience.
• 🎨 Modern UI/UX: Built entirely with Jetpack Compose and Material 3, featuring a sleek "3D" aesthetic, smooth transitions, and Lottie animations.

---

🛠 Technical Stack

Core

• Language: Kotlin
• UI Framework: Jetpack Compose (Material 3)
• Architecture: MVVM (Model-View-ViewModel) + Repository Pattern
• Dependency Injection: Manual injection via ViewModel Factories

Libraries & APIs

• AI/ML: Google Generative AI SDK (Gemini API)
• Camera: CameraX (for high-performance image capture)
• Database: Room Persistence Library (Local SQLite caching)
• Image Loading: Coil (Async image loading for meal logs)
• Animations: Lottie for Compose
• Networking: Kotlin Serialization & Coroutines

---

🏗 Architecture Overview

MacroSnap follows the Clean Architecture principles to ensure scalability and maintainability:

1. Data Layer:
   • MealRepository & AuthRepository act as the single source of truth.
   • MealDao for local persistence.
   • GeminiService for remote AI analysis.
2. Domain Layer: Clean data models (MealAnalysis, MealEntity).
3. UI Layer:
   • ViewModels: State management using StateFlow and collectAsStateWithLifecycle.
   • Composables: Modular and reusable UI components.

---

⚙️ Technical Highlights

AI Prompt Engineering

The app utilizes structured prompt engineering to ensure the Gemini model returns data in a strictly valid JSON format. This allows the application to deserialize AI responses directly into Kotlin data classes using kotlinx.serialization.

CameraX Integration

Custom implementation of the CameraX lifecycle-aware component to capture high-resolution bitmaps which are then optimized/compressed before being sent for AI analysis, ensuring low latency and reduced data usage.

Modern Auth Flow

MacroSnap adopts the latest Credential Manager API, providing a unified sign-in interface that supports Passkeys, Google Sign-In, and traditional passwords, future-proofing the security layer.

---

🛠 Installation & Setup

1. Clone the repository:

   git clone https://github.com/yourusername/MacroSnap.git

2. Obtain a Gemini API Key from Google AI Studio.
3. Add your API key to the local.properties file:

   GEMINI_API_KEY=your_api_key_here

4. Build and run the project using Android Studio Ladybug or later.

---

👨‍💻 Author

Sujoy Dutta -

Project developed as a showcase of modern Android development practices, AI integration, and high-quality UI design.