Food For Health 🥗

A Flutter mobile app that helps people make safer food choices based on their health conditions. Users scan a product's barcode, and the app combines the product's ingredients with the user's chronic conditions to tell them — with the help of generative AI — whether the product is suitable for them, along with a short explanation.

🔌 Backend (ASP.NET Core API): FoodForHealthAPI

---

What it does

• Account & session — register and log in with a phone number; the JWT returned by the API is stored securely and decoded to keep the user signed in across launches.
• Health profile — users record physical information (height, weight, gender, birth date) and their chronic conditions.
• Barcode scanning — scan any product barcode with the device camera.
• AI food evaluation — the scanned product's ingredients and the user's conditions are sent to Google's Gemini model, which returns a clear suitable / not-suitable verdict and a short health note.
• AI chat — a built-in assistant for general nutrition and health questions.

Tech stack

• Flutter (Dart, SDK 3.6+)
• flutter_bloc — Cubit-based state management
• google_generative_ai — Gemini integration for evaluation and chat
• qr_code_scanner — barcode/QR scanning
• http — REST communication with the backend
• flutter_secure_storage + jwt_decoder — secure token storage and session handling
• convex_bottom_bar, quickalert, flutter_markdown, font_awesome_flutter, intl — UI and formatting

Architecture

The project follows a feature-first structure with a shared core layer, and uses the Cubit pattern (view / view_model separation) for state management.

lib/
├── core/                     # Shared building blocks
│   ├── constants/            # API config, colors, prompts, app info
│   ├── general_functions/    # Reusable helpers
│   ├── general_widgets/      # Shared widgets, validators, storage/JWT helpers
│   └── models/               # Data models
└── features/                 # Self-contained features
    ├── splash/
    ├── login/
    ├── register/
    ├── bottom_nav_bar/
    ├── barcode_scan/
    ├── food_evaluation/      # view + view_model (Cubit) + state
    ├── chat_with_ai/
    └── profile/

Each feature keeps its UI (view) separate from its logic (view_model), and screens that have non-trivial state expose a Cubit with an explicit state class.

Engineering highlights

• Feature-first structure — each screen is a self-contained feature with its own view, view-model and state, which keeps the codebase easy to navigate and extend.
• Single networking layer — all REST calls go through one ApiService (core/services) that centralizes the base URL, JSON headers, the JWT bearer token and a shared timeout, so feature view-models stay focused on their own logic instead of repeating HTTP wiring.
• Reusable widgets — shared UI pieces (text fields, validators, buttons, cards, alerts) live in core/general_widgets and are composed across features to avoid duplication.
• Predictable state — Cubits emit explicit loading / loaded / error states, giving the UI a single source of truth to react to.
• Secure session handling — the JWT and cached user are stored with flutter_secure_storage, and the token is attached automatically to every authenticated request.

Screenshots

Login	Register	Barcode scan	Profile

Suitable product	Unsuitable product	Chat with AI

Getting started

Prerequisites

• Flutter SDK (3.6 or newer)
• A running instance of the Food For Health API
• A Google AI Studio API key for the Gemini features

Configuration

Open lib/core/constants/api_constants.dart and set the two values:

// API base URL.
// - Android emulator:  http://10.0.2.2:<port>
// - iOS simulator:     http://localhost:<port>
// - Physical device:   http://<your-computer-LAN-IP>:<port>
static const String apiBaseUrl = "http://10.0.2.2:5016";

// Your Google AI Studio (Gemini) API key.
static const String geminiAPIKey = "";

Android blocks plain HTTP (cleartext) traffic by default. For local development against an HTTP API, usesCleartextTraffic is enabled in the Android manifest.

Run

flutter pub get
flutter run

License

Released under the MIT License.