RAG Store Assistant

Homepage

Introduction

RAG Store Assistant is an intelligent chatbot system designed to assist users in searching for products, answering FAQs, and providing personalized recommendations. Built on the RAG (Retrieval-Augmented Generation) model, the system combines Hybrid Search (Semantic Search + Keyword Search) for efficient information retrieval and Gemini-1.5-flash for natural language generation. The backend is powered by FastAPI, while the frontend is built using Node.js. The system is containerized using Docker for easy deployment and scalability.

This project is the final assignment for the course CS311.P11, developed by the following team members:

• Đinh Thiên Ân - 22520010
• Huỳnh Trọng Nghĩa - 22520003
• Lê Trần Gia Bảo - 22520105

Features

• Product Search: Find products based on user queries (e.g., name, color, size).
• FAQ Support: Automatically answer common questions about policies, shipping, and returns.
• Order Tracking: Provide real-time updates on order status.
• Hybrid Search: Combines Semantic Search and Keyword Search for accurate and flexible retrieval.
• Natural Language Generation: Uses Gemini-1.5-flash to generate human-like responses.

Installation

Prerequisites

• Python 3.8+
• Node.js 16+
• Docker (optional)

Steps

1. Clone the repository:

   git clone https://github.com/your-username/rag-store-assistant.git
   cd rag-store-assistant

2. Install backend dependencies:

   pip install -r requirements.txt
   

3. Install frontend dependencies:

   cd frontend
   npm install

4. Set up environment variables:

   MONGODB_URI=your_mongodb_connection_string
   GEMINI_API_KEY=your_gemini_api_key
   

Usage

Backend

To start the backend server, run:

uvicorn app:app --reload

Frontend

To build and run the frontend, navigate to the frontend directory and run:

npm run build
npm run dev

The frontend will be available at http://localhost:5127.

Project Structure

``` bash
rag-store-assistant/
├── backend/               # Backend code (FastAPI)
│   ├── app/               # Main application logic
│   ├── models/            # Data models
│   ├── utils/             # Utility functions
│   └── requirements.txt   # Python dependencies
├── frontend/              # Frontend code (Node.js)
│   ├── public/            # Static assets
│   ├── src/               # React components
│   └── package.json       # Node.js dependencies
├── docker-compose.yml     # Docker configuration
└── README.md              # Project documentation

License

This project is licensed under the MIT License. See the LICENSE file for details.