Uploaded April 2026 | Updated September 2026, 2 weeks ago
The core argument: AI systems need more than top-K chunks. They need structured context about entities, relationships, permissions, authorship, provenance, and history. GraphRAG combines vector search with graph traversal so retrieval can start semantically, then expand through meaningful relationships. This makes answers more accurate, easier to debug, and more explainable. Emil Eifrem, CEO of Neo4J, explains why graph databases are becoming newly important in AI systems. The conversation covers Neo4j’s origin, GraphRAG, knowledge graphs, agent memory, and why future AI applications may need graph-shaped context layers.
Timestamps
00:00:00 Why graphs matter now
00:04:36 The origin of Neo4j
00:09:44 Graph databases in plain English
00:15:20 Fraud, identity, and real-time context
00:21:28 Knowledge graphs meet RAG
00:28:33 GraphRAG and agent memory
00:35:32 Modeling, tooling, and developer experience
00:42:05 The future graph-shaped internet
The core argument: AI systems need more than top-K chunks. They need structured context about entities, relationships, permissions, authorship, provenance, and history. GraphRAG combines vector search with graph traversal so retrieval can start semantically, then expand through meaningful relationships. This makes answers more accurate, easier to debug, and more explainable. Emil Eifrem, CEO of Neo4J, explains why graph databases are becoming newly important in AI systems. The conversation covers Neo4j’s origin, GraphRAG, knowledge graphs, agent memory, and why future AI applications may need graph-shaped context layers.
Timestamps
00:00:00 Why graphs matter now
00:04:36 The origin of Neo4j
00:09:44 Graph databases in plain English
00:15:20 Fraud, identity, and real-time context
00:21:28 Knowledge graphs meet RAG
00:28:33 GraphRAG and agent memory
00:35:32 Modeling, tooling, and developer experience
00:42:05 The future graph-shaped internet
