Essential GraphRAG: Knowledge Graph-Enhanced RAG w/ Tomaž Bratanič & Oskar Hane

We'll explore how Knowledge Graphs can significantly improve Retrieval-Augmented Generation (RAG) systems. Tomaž Bratanič and Oskar Hane will share practical methods for building more accurate, traceable, and robust RAG applications by integrating structured data from KGs with traditional unstructured data retrieval.

What You'll Learn:

• GraphRAG Architecture: Integration of knowledge graphs with RAG pipelines to handle both structured queries ("How many active contracts?") and semantic retrieval
• Hybrid Search Strategy: Combining vector similarity, full-text search, and graph queries for comprehensive information retrieval
• Text2Cypher Generation: Converting natural language to Cypher queries - enabling dynamic graph querying without hardcoded templates
• Agentic RAG Systems: Multi-agent architectures with intelligent routing between vector search, graph queries, and specialized tools
• Entity Resolution: Systematic approaches to merging different representations of the same real-world entity
• Microsoft's GraphRAG: Community detection and hierarchical summarization for both local and global search capabilities
• Structured Data Extraction: Using LLMs to extract entities and relationships from unstructured text into knowledge graphs

Agenda:

• 12:00 - 12:10 PM: Welcome and community updates
• 12:10 - 12:50 PM: Main presentation with examples
• 12:50 - 1:20 PM: Q&A and discussion
• 1:20 - 1:30 PM: Manning Publications raffle and book giveaways

Who Should Attend: Developers and data practitioners working with RAG systems, knowledge graphs, or looking to build more reliable AI applications with structured + unstructured data.
Format: Virtual lunch & learn (12:00-1:30 PM EST Friday) - European-friendly timing. Grab your lunch and join us!