Building an Explainable GraphRAG System: From Clinical Notes to Knowledge Graphs with Neo4j

Session track: AI Engineering

Session time: 12:30 pm - 01:00 pm UTC November 12

Session description:

Show how to build a complete GraphRAG pipeline that converts unstructured clinical notes into an explainable Neo4j knowledge graph using LLMs. Cover entity extraction, graph modelling, temporal relationships, Cypher queries, GraphRAG retrieval, and lessons learned. Demonstrate how graph-based reasoning provides transparent, evidence-backed answers and discuss how the same architecture can be applied beyond healthcare.