Uploaded July 2026 | Updated September 2026, 19 hours ago

---

Neo4j CTO Philip Rathle sits down with Michael Grinich at AI Engineer World's Fair 2026 to explain why more than 70% of Neo4j's new business last quarter came from AI, not traditional graph database queries.

They cover:
• Why knowledge graphs are becoming the "left brain" to LLMs' right brain, especially for questions where 100% accuracy is non-negotiable (fraud, compliance, drug discovery)
• How companies with dozens of siloed systems, including one with 50 separate ERP systems, use graphs to unify context for AI
• The shift from vector-only RAG to graph RAG, and why that matters for multi-hop reasoning
• Why chaining AI agents together compounds errors, and what a supervisor layer looks like in practice
• How Neo4j uses AI internally to speed up engineer onboarding and troubleshoot production issues

Watch for Philip's take on where knowledge graphs go next as agentic AI scales.

Neo4j: neo4j.com
WorkOS: workos.com