Graph and Lakehouse, Friends at Last
This title could be clearer and more informative.Try out Clickbait Shieldfor free (5 uses left this month).
Neo4j Virtual Graph is a new capability that overlays a property graph model on top of existing Databricks lakehouse data without moving it. Foreign keys and join tables become explicit, named relationships queryable via Cypher, which is then deterministically translated to SQL and executed on Databricks compute under Unity Catalog governance. The post walks through setting up Virtual Graph on the TPC-H benchmark dataset, covering schema preparation, composite key workarounds, data source connection, AI-assisted graph model generation, and several progressively complex Cypher query patterns — including star schema navigation, bipartite intersection, parametric lookups for agentic workflows, and multi-role supplier scoring — comparing each to its SQL equivalent to illustrate the semantic clarity gains.
