Retrieval-Augmented Generation (RAG) is the cornerstone of many Gen AI applications – from AI assistants to intelligent search to custom copilots. GraphRAG adds knowledge graphs to traditional RAG to provide structured relationships and deeper context, allowing it to connected ideas together instead of individual documents. In this workshop you will learn how to use AllyCat, an open source application from the AI Alliance, to build your own Agent using using real-world content scraped from websites. We will walk through the two halves of AllyCat:
1) Data: Crawling and scraping website data;
2) Q&A: Using your own data, knowledge data and an LLM to generate context-aware answers.
By the end, you will have created a working GraphRAG prototype powered by content you extracted yourself. This workshop is ideal for anyone looking to get their hands dirty with open source tools like Docling, Data Prep Kit, Milvus, Neo4j, Granite embedding models and Llama LLMs. No prior experience required.
