Built an AI-powered “Chat with Website” application using Next.js, JavaScript, and Tailwind CSS that allows users to enter any website URL and interact with its content through a smart chat interface. The system uses RAG (Retrieval-Augmented Generation) with Google Gemini API to understand and answer questions based on the website data.
The application scrapes and processes webpage content, converts it into embeddings, and stores them in Upstash vector database for fast semantic retrieval. When a user asks a question, the most relevant content chunks are fetched and sent to the AI model to generate accurate, context-aware responses.
- 🌐 Chat with any website using its URL
- 🧠 RAG-based semantic search over webpage content
- ⚡ Real-time AI responses using Gemini API
- 📦 Fast vector storage with Upstash
- 💬 Interactive chat UI with history support
- 📚 Turns any website into a Q&A knowledge base
- ⏱️ Saves time by avoiding manual reading of long pages
- 🔍 Improves research efficiency for students & developers
- 🧠 Provides accurate answers grounded in real website data
- 🚀 Useful for documentation, blogs, and research papers
- Next.js
- JavaScript
- Tailwind CSS
- Google API
- Upstash Vector DB
- RAG pipeline (embedding + retrieval + generation)
This project transforms any static website into an interactive AI knowledge assistant, making web content searchable, conversational, and instantly accessible 🌐🤖