User Guide#

Learn how to use euclid_rag for querying Euclid space mission documents.

Overview#

Euclid RAG is an open-source Retrieval-Augmented Generation (RAG) system designed to provide efficient document retrieval and knowledge augmentation for the Euclid scientific community. The project integrates local Large Language Models (LLMs) with a vector database to retrieve, process, and generate relevant scientific information.

Key Features#

  • Local deployment without API-based LLM dependencies

  • Document retrieval strategies tailored to Euclid’s scientific workflows

  • Streamlit-based user interface for easy interaction

  • FAISS vector store for efficient semantic search

  • Multiple document types including publications and DPDD documents

  • Docker support for containerized deployment

  • Potential agentic capabilities for automated knowledge retrieval

Quick Start#

Get up and running with euclid_rag in four steps:

  1. Install euclid_rag using your preferred method

  2. Configure your system and vector store settings

  3. Ingest documents into the vector store

  4. Run the chatbot interface

Project Origins#

This project was initially forked from the Rubin Observatory’s Rubin RAG system. While developed in consultation with Rubin developers, Euclid RAG is evolving in a different direction to meet the specific needs of the Euclid collaboration.

Key Differences:

  • Focus on local deployment without API dependencies

  • Different document retrieval strategies for Euclid workflows

  • Potential agentic capabilities for automated processing

Next Steps#

User Guide: