LustBot - E-Commerce Chatbot

AI-powered sales and customer service chatbot for MyLastShop, built with Pydantic AI and FastAPI.

Features

• AI Sales Agent ("לסטי"): Hebrew-speaking chatbot powered by Google Gemini
• Knowledge Base: MongoDB Atlas vector search for product information
• Order Management: Google Sheets integration for order storage
• Human Escalation: Automatic detection and handoff to human support
• Session Memory: Conversation history persistence
• Modern Chat UI: RTL Hebrew support with responsive design

Quick Start

Prerequisites

• Python 3.11+
• Google Gemini API key
• OpenAI API key (for embeddings)
• MongoDB Atlas cluster with vector search enabled
• Google Sheets service account credentials

Installation

1. Clone and setup

cd LustBot-claude\ code
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate
pip install -r backend/requirements.txt

2. Configure environment

# Edit .env file with your credentials
cp .env.example .env

3. Add Google Sheets credentials Place your credentials.json file in the project root.

4. Run the backend

cd backend
uvicorn app.main:app --reload --port 8000

5. Open the frontend Open frontend/index.html in your browser, or serve it:

# Using Python's built-in server
cd frontend
python -m http.server 3000

Then visit http://localhost:3000

Docker Deployment

docker-compose up -d

Access the chat at http://localhost

Project Structure

ecommerce-chatbot/
├── backend/
│   ├── app/
│   │   ├── agents/         # Pydantic AI agent
│   │   ├── tools/          # Vector store, Google Sheets, escalation
│   │   ├── models/         # Pydantic models
│   │   ├── services/       # Memory, MongoDB, embeddings
│   │   ├── routers/        # API endpoints
│   │   └── main.py         # FastAPI entry point
│   ├── requirements.txt
│   └── Dockerfile
├── frontend/
│   ├── index.html
│   ├── css/styles.css
│   └── js/
│       ├── api.js
│       └── chat.js
├── docker-compose.yml
├── nginx.conf
└── .env

API Endpoints

Method	Endpoint	Description
POST	/api/chat	Send message and get response
GET	/api/history/{session_id}	Get conversation history
DELETE	/api/history/{session_id}	Clear session history
GET	/api/admin/health	Health check
GET	/api/admin/sessions	List active sessions
GET	/api/admin/escalations	List escalation history

Configuration

Environment Variables

Variable	Description
GOOGLE_API_KEY	Google Gemini API key
OPENAI_API_KEY	OpenAI API key for embeddings
MONGODB_URI	MongoDB Atlas connection string
MONGODB_DATABASE	Database name
MONGODB_COLLECTION	Collection name for vector store
MONGODB_VECTOR_INDEX	Vector search index name
GOOGLE_SHEETS_CREDENTIALS_PATH	Path to service account JSON
GOOGLE_SHEETS_SPREADSHEET_ID	Google Sheets ID
GOOGLE_SHEETS_SHEET_NAME	Sheet name for orders

MongoDB Vector Search Setup

Ensure your MongoDB collection has a vector search index:

{
  "mappings": {
    "dynamic": true,
    "fields": {
      "embedding": {
        "dimensions": 1536,
        "similarity": "cosine",
        "type": "knnVector"
      }
    }
  }
}

Escalation Keywords

The bot automatically detects these Hebrew keywords and offers human support:

• אדם, נציג, בן אדם, מנהל
• עזור לי, עזרה, תלונה
• החזר, זיכוי, דחוף
• And more...

License

Private - MyLastShop