A versatile and intelligent AI chatbot built with LangGraph that can be integrated into any e-commerce platform to provide customers with natural language product discovery and assistance.
ShopAssist is a powerful conversational agent that leverages Google's Gemini model, LangGraph for orchestration, and vector search to deliver a seamless shopping experience. The system can understand complex queries, maintain context throughout a conversation, and perform targeted searches using tool calling capabilities.
- 🧠 LangGraph Agent Architecture - Advanced conversational flow with state management
- 🔧 Tool Calling Integration - Agent can call specialized tools to fulfill user requests
- 🔍 Vector Search Capabilities - Semantic product search using MongoDB Atlas Vector Database
- 💬 Context-Aware Conversations - Maintains conversation history for better responses
- 🔄 Flexible Integration - Designed to work with any e-commerce platform
- 🛒 Product Discovery - Helps users find products through natural language
- 📱 Responsive Design - Works seamlessly on both desktop and mobile devices
- LangGraph - For creating a conversational graph with stateful agents
- Google Generative AI - Utilizing Gemini 1.5 Flash model for natural language understanding
- Vector Database - MongoDB Atlas with vector search capabilities
- Node.js with Express
- TypeScript for type safety
- MongoDB for product and conversation storage
- React with TypeScript
- Tailwind CSS for styling
- Vite as the build tool
The system is built around a LangGraph agent that can:
- Understand user intentions through natural language processing
- Call specialized tools based on the conversation context
- Perform vector searches to find semantically similar products
- Track conversation state to maintain context
- Generate helpful responses based on product information
- Node.js (v18+ recommended)
- MongoDB Atlas account with Vector Search enabled
- Google AI Studio API key
- Clone this repository
git clone <repository-url>
cd ecom-chat- server directory
cd server
npm install
npm run dev- client directory
cd client
npm install
npm run dev