Strategy One (March 2026) opens Mosaic MCP server by default for AI agents

The walkthrough builds on two earlier posts in the series that used the same Customer 360 pattern with different semantic layers: Stardog's ontology-based layer and Denodo's data virtualization platform. This third post swaps in Strategy Mosaic, whose dimensional model composes attributes, hierarchies, facts, and metrics through Strategy's metric expression syntax. The shared C360 schema is adapted from Stardog's C360 Knowledge Kit, so a reader following the series sees the same entities and keys answered by a different semantic layer.

Mosaic does not federate rows across engines at query time. It plans one query per source, pushes native SQL down, and joins the intermediate result sets in the Mosaic engine on shared attributes declared through model linking. For the representative auto-enrollment question, Amazon Aurora runs a transactional profile join while Amazon Redshift runs the fact aggregation, and Mosaic combines the results on the shared customer ID. The body notes that domain ownership follows the model boundary, with the CRM team owning Customer and the data system team owning Purchase, and neither hands over its source.

A notable detail is that Mosaic serves four native protocols from one definition: MCP for AI agents, JDBC/ODBC over Trino for SQL clients, XMLA for Power BI and Excel, and REST for custom applications. The body also describes a choice between live pushdown and in-memory execution, noting that agent traffic is repetitive and bursty, so serving from memory can turn variable per-query warehouse spend into fixed Mosaic capacity.