GraphGeeks Talk: Smarter RAG Starts with Cleaner Graphs

Building LLM apps is easy, but getting high accuracy is hard. Most developers use vector databases for RAG, but adding knowledge graphs can deliver dramatically better results - IF you solve the duplicate entity problem first. In this talk, Dr. Clair Sullivan reveals how Entity-Resolved Knowledge Graphs (ERKGs) transform RAG performance by eliminating the confusion caused by duplicate entities across data sources. She demonstrates real-world examples showing measurable accuracy improvements when LLMs work with clean, consolidated entity data instead of messy duplicates. You'll find out: - Why duplicate entities kill RAG accuracy (and how to spot them) - Advanced entity resolution techniques that go beyond string matching - Practical steps to clean your knowledge graphs for better LLM results Perfect for data scientists, AI engineers, and developers building RAG systems who want to move beyond basic implementations to achieve better accuracy.