This project provides an AI chatbot for analyzing and extracting information from data via a conversational interface, built using Langchain. It targets developers and users who need to query documents and data programmatically, offering interactive UIs for data ingestion and chat interaction with adjustable parameters.

How It Works

The chatbot leverages Langchain for orchestrating complex LLM workflows. It utilizes Pinecone as a vector database for efficient similarity search on ingested data, enabling context-aware responses. OpenAI's API is used for the core language model capabilities. The architecture supports interactive UIs for both file ingestion and chat, allowing users to directly interact with and configure the chatbot's behavior.

Quick Start & Requirements

• Install: Clone the repository, create a .env file from example.env with API keys for OpenAI and Pinecone, then run poetry install, poetry shell, and python3 startup.py.
• Prerequisites: Git, Pinecone account, OpenAI API key (requires billing).
• Setup: Estimated setup time is minimal, but requires obtaining API keys.
• Docs: Discord for help.

Highlighted Details

• Interactive Ingestion UI for files.
• Chat UI with configurable parameters (source, temperature, vector_k).
• Python port of a popular GPT4 & LangChain chatbot for PDF documents.

Maintenance & Community

• Developed by Haste171, inspired by Mayo.
• Community support available via Discord.

Licensing & Compatibility

• MIT License. Permissive for commercial use and integration with closed-source projects.

Limitations & Caveats

Currently supports limited file types, with plans to expand compatibility to more formats and offline models. The reliance on OpenAI's API means it requires a paid account and internet connectivity.