This project provides a FastAPI application enabling conversational interaction with PDF documents. It targets researchers and users needing to quickly extract information from papers, offering responses with source citations and page numbers. The core benefit is an intuitive interface for querying research papers via uploaded PDFs or online links.

How It Works

The application processes PDFs by extracting text, generating embeddings, and leveraging the OpenAI API for question answering. It stores embeddings locally using Redis. This approach allows for efficient semantic search within the document content, enabling the LLM to pinpoint relevant information and provide accurate, sourced answers.

Quick Start & Requirements

• Install via pip install -r requirements.txt after cloning the repository.
• Requires Redis for local embedding storage.
• An OpenAI API key is mandatory, set as the OPENAI_API_KEY environment variable.
• Demo available at: https://www.dara.chat

Highlighted Details

• Conversational interface for PDF documents.
• Generates responses with source citations and page numbers.
• Supports online PDF links and local uploads.

Maintenance & Community

This repository is no longer actively maintained. A more advanced version is available at chat.portal.so, with plans for future open-sourcing.

Licensing & Compatibility

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

Limitations & Caveats

The project is explicitly stated as no longer actively maintained. Users should refer to the successor project for current features and support.