mise - AI Culinary Assistant

A personal AI sous-chef that discovers recipes, respects your dietary needs, and helps you plan the grocery run.

Key Features

• Conversational AI Chat - Describe what you want to cook or what's in your fridge. The AI finds recipes, handles substitutions, and adjusts for dietary needs.
• Fridge Inventory - Log what you have on hand. The AI prioritizes recipes using your available ingredients and flags items expiring soon.
• Taste Profile - Save allergies, dietary restrictions, favorite and disliked ingredients. Every recommendation is personalized automatically.
• Recipe Collection - Save recipes to your personal cookbook. Search and filter by cuisine, difficulty, or cook time.
• Smart Shopping Lists - Add a recipe's ingredients to your shopping list. Items already in your fridge are flagged so you don't buy duplicates.

Tech Stack

Layer	Technology
Frontend	React 19, Styled Components, Framer Motion
Backend	Node.js, Express, Helmet, Rate Limiting, Winston Logger
Database	MongoDB with Mongoose ODM
AI	Together.ai API (LLM with fridge-aware context)
Auth	JWT + bcrypt
Validation	express-validator

Getting Started

Prerequisites

• Node.js (v16+)
• MongoDB (local or MongoDB Atlas)
• Together.ai API key

Installation

1. Clone the repository:

   git clone https://github.com/junho5400/mise.git
   cd mise

2. Install dependencies:

   npm install
   cd client && npm install && cd ..

3. Set up environment variables:

   Create a .env file in the root directory (see .env.example):

   PORT=3000
   MONGODB_URI=your_mongodb_connection_string
   JWT_SECRET=your_jwt_secret
   TOGETHER_API_KEY=your_together_ai_api_key
   CORS_ORIGIN=http://localhost:3001
   LLM_MODEL=meta-llama/Llama-3.3-70B-Instruct-Turbo-Free
   LLM_RATE_LIMIT=5
   LOG_LEVEL=info

4. Start the development servers:

   # Backend (port 3000)
   npm run dev
   
   # Frontend (separate terminal)
   cd client && npm start

API Endpoints

Authentication

Method	Endpoint	Description
POST	/auth/register	Register a new user
POST	/auth/login	Log in
GET	/auth/me	Get current user profile

Chat

Method	Endpoint	Description
POST	/api/chat	AI conversation (uses preferences + fridge contents as context)

User

Method	Endpoint	Description
POST	/users/me/recipes/:recipeId	Save recipe
DELETE	/users/me/recipes/:recipeId	Remove saved recipe
GET	/users/me/recipes	Get saved recipes
PUT	/users/me/preferences	Update dietary preferences

Fridge

Method	Endpoint	Description
GET	/users/me/fridge	Get all fridge items
POST	/users/me/fridge	Add a fridge item
PUT	/users/me/fridge/:itemId	Update a fridge item
DELETE	/users/me/fridge/:itemId	Remove a fridge item
DELETE	/users/me/fridge/expired	Clear all expired items

Shopping List

Method	Endpoint	Description
GET	/shopping-list	Get shopping list
POST	/shopping-list/recipe/:recipeId	Add recipe ingredients (flags items already in fridge)

Project Structure

├── src/                    # Backend
│   ├── index.js            # Server entry point
│   ├── config/             # Database, env validation, logger
│   ├── controllers/        # Auth, recipe, user, fridge, shopping list
│   ├── models/             # Mongoose schemas (User with fridge, Recipe, ShoppingList)
│   ├── routes/             # Express route definitions
│   ├── middleware/          # Auth, validation, error handling
│   └── services/           # LLM integration (fridge-aware prompting)
├── client/src/             # Frontend (React)
│   ├── EnhancedApp.jsx     # Root component with routing
│   ├── context/            # Auth state management
│   └── components/         # UI components
│       ├── EnhancedHome.jsx         # Landing page
│       ├── EnhancedChatInterface.jsx # AI chat (two-panel layout)
│       ├── Fridge.jsx               # Fridge inventory
│       ├── EnhancedPreferences.jsx  # Taste profile
│       ├── EnhancedSavedRecipes.jsx # Recipe collection
│       └── ShoppingList.jsx         # Shopping list
└── .env.example            # Environment variable template

Deployment

1. Build the React frontend:

   cd client && npm run build

2. Deploy to your preferred hosting platform

3. Set all required environment variables on your hosting platform

Author

Junho Hong

License

This project is licensed under the MIT License.