A multi-agent customer service chatbot for an e-commerce platform that can filter messages, classify user intent, guide users through ordering, retrieve product information, and offer personalized recommendations.

The system follows a modular pipeline structure:

User ->
Guard Agent ->
Classification Agent ->
   -> Order Taking Agent
   -> Details Agent
   -> Recommendation Agent
Back to User

• Guard Agent: Filters out irrelevant or inappropriate messages before processing
• Classification Agent: Accurately classifies user intent and routes queries to specialized agents
• Order Taking Agent: Guides users through the ordering process using step-by-step reasoning
• Details Agent: Retrieves and responds with detailed product information (RAG System)
• Recommendation Agent: Offers personalized product recommendations based on user preferences

• Python 3.9+
• FastAPI
• SQLAlchemy
• PostgreSQL
• FAISS for vector search
• Hugging Face Transformers

• React
• React Router
• Axios

├── backend/             # Backend code
│   ├── app/            # FastAPI application
│   │   ├── agents/     # Agent implementations
│   │   ├── database/   # Database models and connection
│   │   ├── faiss/      # Vector search implementation
│   │   ├── routes/     # API endpoints
│   │   ├── services/   # Business logic services
│   │   └── utils/      # Utility functions
│   ├── main.py         # Entry point for the backend
│   └── requirements.txt # Python dependencies
│
└── frontend/           # React frontend
    ├── public/         # Static files
    ├── src/            # React source code
    │   ├── components/ # Reusable UI components
    │   ├── pages/      # Page components
    │   └── services/   # API services
    └── package.json    # Node.js dependencies

• Python 3.9+
• Node.js and npm
• PostgreSQL

1. Clone the repository:

git clone https://github.com/yourusername/customer-service-chatbot.git
cd customer-service-chatbot

2. Create and activate a virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install Python dependencies:

cd backend
pip install -r requirements.txt

4. Set up environment variables by creating a .env file in the backend directory:

DATABASE_URL=postgresql://username:password@localhost:5432/ecommerce_chatbot

1. Create a PostgreSQL database:

createdb ecommerce_chatbot

2. Run database migrations (if applicable)

You can use the provided scripts to start the application:

# Start both backend and frontend
./run.sh

# Start only the backend
./run.sh backend

# Start only the frontend
./run.sh frontend

# Build the frontend
./run.sh build

# Start both backend and frontend
run.bat

# Start only the backend
run.bat backend

# Start only the frontend
run.bat frontend

# Build the frontend
run.bat build

Alternatively, you can start the services manually:

cd backend
python -m uvicorn app.main:app --reload --host 0.0.0.0 --port 8000

The API will be available at http://localhost:8000

cd frontend
npm install
npm run dev

The frontend will be available at http://localhost:5173

1. Build the React frontend:

cd frontend
npm run build

2. Start the production server:

cd backend
python -m uvicorn app.main:app

The application will serve both the API and the frontend from http://localhost:8000

• /api/chat/ - Process chat messages
• /api/order/ - Handle order operations
• /api/recommend/ - Get product recommendations
• /api/faq/ - Access frequently asked questions

MIT License