Uploaded June 2026 | Updated September 2026, 2 weeks ago
In Part 2, Rick Houlihan demonstrates Unified Model Theory on Oracle AI Database 26ai: one canonical, normalized schema, projected into every shape its consumers need. From a single restaurant model we'll serve the point-of-sale app a document, analytics a table, recommendations a graph, and search a vector index — over the same tables, one optimizer, one transaction, no copies and no eventual consistency.
We'll build JSON Relational Duality Views that read as documents and decompose writes back to rows (with per-field governance and lock-free ETAG concurrency), see where a duality view is the wrong tool (the hot cart that wants a single collection), and settle the tree-versus-graph question — a materialized path with a multivalue index for hierarchies, a real property graph for true traversal.
The payoff isn't a faster query; it's the ORM layer, the CDC pipelines, the sagas, and the consistency windows that simply stop existing.
Part 2 of 2! Part 1 goes into the theory and discusses the “knobs to turn”.
In Part 2, Rick Houlihan demonstrates Unified Model Theory on Oracle AI Database 26ai: one canonical, normalized schema, projected into every shape its consumers need. From a single restaurant model we'll serve the point-of-sale app a document, analytics a table, recommendations a graph, and search a vector index — over the same tables, one optimizer, one transaction, no copies and no eventual consistency.
We'll build JSON Relational Duality Views that read as documents and decompose writes back to rows (with per-field governance and lock-free ETAG concurrency), see where a duality view is the wrong tool (the hot cart that wants a single collection), and settle the tree-versus-graph question — a materialized path with a multivalue index for hierarchies, a real property graph for true traversal.
The payoff isn't a faster query; it's the ORM layer, the CDC pipelines, the sagas, and the consistency windows that simply stop existing.
Part 2 of 2! Part 1 goes into the theory and discusses the “knobs to turn”.
