This project provides a Python-based chatbot application that allows users to interact with any website by providing its URL. Targeting developers and AI enthusiasts, it simplifies the process of building a Retrieval-Augmented Generation (RAG) system with a user-friendly Streamlit GUI, enabling efficient information extraction and LLM-powered querying.

How It Works

The application employs a Retrieval-Augmented Generation (RAG) approach. It first scrapes and vectorizes text content from a given website URL. When a user queries the chatbot, the system retrieves the most relevant text chunks from the vectorized data based on semantic similarity to the query. This retrieved context is then prepended to the user's prompt, which is subsequently sent to a chosen Large Language Model (LLM) like GPT-4, Mistral, or Llama2 for generating an informed response.

Quick Start & Requirements

• Install dependencies: pip install -r requirements.txt
• Set environment variable: OPENAI_API_KEY=[your-openai-api-key]
• Run the Streamlit app: streamlit run app.py
• Requires Python.

Highlighted Details

• Leverages LangChain 0.1.0 for website interaction and RAG.
• Supports multiple LLMs including GPT-4, Mistral, Llama2, and ollama.
• Features an intuitive GUI built with Streamlit.
• Primarily serves as supporting material for a YouTube tutorial.

Maintenance & Community

The repository is intended as supporting material for a YouTube video tutorial. Pull requests are only accepted for bug or typo fixes.

Licensing & Compatibility

Licensed under the MIT License. For educational and research purposes; users must comply with API terms of use.

Limitations & Caveats

The project is primarily educational and may not be production-ready. Pull requests are restricted, limiting community contributions beyond bug fixes.