At Oracle AI World in Las Vegas, Google Cloud and Oracle announced an expansion of their partnership across the AI stack—AI compute and infrastructure, enterprise data, and agentic AI—by deepening integration between Oracle AI Database@Google Cloud, Google’s Agentic Data Cloud, and Gemini Enterprise. The aim is to combine Oracle’s mission-critical operational data and applications with Google Cloud‘s data and AI ecosystem, including BigQuery, Gemini Enterprise, and Gemini models, so enterprises can build autonomous agentic workflows instead of static analytics. The architecture gives agents direct, secure, real-time access to foundational enterprise data.

The post lays out several building blocks for production-grade agentic AI on Oracle AI Database@Google Cloud. The Remote Model Context Protocol (MCP) Server for Oracle workloads, powered by the IAM-governed ExecuteSQL Data API, is now generally available. It provides a unified interface for AI agents, Google Cloud services, and enterprise applications to run SQL against Oracle databases, with credentials resolved automatically through Secret Manager and Google Cloud IAM enforcement, removing the need for custom natural-language-to-SQL middleware, bastion VMs, or plaintext passwords. The Knowledge Catalog integration, in preview, unifies metadata discovery and governance across Oracle databases and Google Agentic Data Cloud by cataloging schemas and business metadata automatically.

A new preview, Gemini connector for Oracle AI Database@Google Cloud, lets business users and developers query Oracle databases in natural language inside the Gemini app; Gemini uses Knowledge Catalog to discover, search, and ground responses in governed Oracle data each user is authorized to access. Also announced: Oracle in Gemini Enterprise, available today, adds conversational analytics, cross-cloud database connectors, and domain-aware agents, including the Oracle AI Database Agent, with more than 135 pre-built operational and developer skills and standardized MCP connectivity with zero plaintext credentials. Gemini frontier reasoning is embedded in Oracle Fusion AI Agent Studio, Fusion Claw, Oracle Fusion Applications, and NetSuite. Partner-ready agentic blueprints are being built with Argano, Automus, Calfus, HCL, IBM, Infosys, LTM, Pythian, TCS, and Tessell.

For data unification, Google Cloud and Oracle offer three architectural choices. The bi-directional borderless Lakehouse and Oracle AI Lakehouse integration uses shared Apache Iceberg tables via a Lakehouse runtime REST catalog integrated through DBMS_CATALOG. A new preview, Oracle Access to Google Cloud borderless Lakehouse, lets Oracle Autonomous AI Database and BigQuery/Managed Spark directly query and execute write/DML operations on shared Iceberg tables in Google Cloud Storage without ETL or replication. OCI GoldenGate provides real-time CDC from Oracle redo logs, streaming sub-second analytical freshness into BigQuery managed tables without touching table blocks. BigQuery query federation with Oracle, in preview, uses EXTERNAL_QUERY for real-time in-place analytics against live Oracle databases with zero data movement.

Finally, two new regions—europe-west12-a (Turin, Italy) and northamerica-south1-b (Queretaro, Mexico)—are live for Oracle AI Database@Google Cloud, and Oracle announced plans to pursue FedRAMP High certification for the service, allowing US government and public sector entities to run regulated workloads on Oracle Exadata Database Service within Google Cloud Assured Workloads. The service is available through Google Cloud Marketplace with existing Google Cloud commitments, unified billing, and Oracle Support Rewards.