2025 ODSC Workshop: Adaptive RAG Systems with Knowledge Graphs: Building Reinforcement-Learning-Driven AI Applications

Welcome to the landing page for the workshop Adaptive RAG Systems with Knowledge Graphs: Building Reinforcement-Learning-Driven AI Applications at the `2025 Open Data Science Conference (ODSC) East.

What to Expect

This workshop intends to provide an introduction to:

• Build a naively simple Graph-based Retrieval-Augmented Generation (RAG) Agent using LLM-generated Cyphers
• Build a useful Graph-based RAG Agent using Fixed-Logic Cyphers
• (Optional) Introduction to using Google's Agent2Agent (A2A) Protocol to connect multiple agents together

All of this can be done without requiring GPU resources or lengthy training times because:

• Graph Databases don't (typically) require GPUs!
• IMPORTANT The training dataset is intentionally small for the purposes of this workshop

Participation Options

There are 2 options available to participate in this workshop:

Option 1: The Easy Way to Participate

If you are just looking to learn about RAG Agents, we will offer several Google Colab Notebooks where you can participate in the workshop without requiring any prerequisite software or configuration. You need a laptop and a browser.

If you choose this option, click COLAB_INSTRUCTIONS.md.

Option 2: A Full Production-like Experience

If you opt for the full experience, you need a Linux/MacOS-based operating system. You should have already installed the software components in the pre-workshop instructions.

If you haven't set anything up yet and still want to choose this option, the link below will have all the software you need to install.

If you choose this option, click LAPTOP_INSTRUCTIONS.md.