Oracle Deepens Google Cloud Integration: Select AI Embraces Gemini Enterprise Agent Platform
Oracle continues to strengthen its multicloud alliance with Google by adding Gemini Enterprise Agent Platform — Google’s rebranded successor to Vertex AI — as a fully supported AI provider for Select AI on Autonomous AI Database Serverless. Announced on October 13, 2026, this integration allows Oracle database professionals to harness Google’s flagship Gemini models directly from within their SQL environment, no custom plumbing required.
For Oracle DBAs and developers already operating on Oracle Database@Google Cloud, this is a significant quality-of-life improvement. It eliminates the need for bespoke middleware or API wrappers to access Gemini models, instead channeling everything through the familiar Select AI interface that many teams have already adopted for natural language to SQL workflows.
What’s New: Gemini Enterprise Agent Platform as a First-Class Select AI Provider
With this update, users can now configure Select AI profiles to specify Gemini Enterprise Agent Platform as their AI provider. Once configured, the full spectrum of Select AI capabilities becomes available through Google’s Gemini models:
- NL2SQL (Natural Language to SQL): Ask questions in plain English and receive executable SQL queries generated by Gemini models, grounded in your database schema.
- Chat: Engage in conversational interactions with your data using Gemini’s advanced language understanding.
- ShowSQL: Inspect the SQL that Select AI generates from your natural language prompts before execution.
- ExplainSQL: Get human-readable explanations of complex SQL statements, powered by Gemini’s reasoning capabilities.
- Embeddings: Generate vector embeddings directly within the database for semantic search, retrieval-augmented generation (RAG), and other AI-driven data operations.
The integration is designed to be project-scoped and region-scoped, meaning customers access exactly the models available in their specific Google Cloud project and region. This approach respects data residency requirements and organizational governance policies — a critical consideration for enterprise deployments.
Authentication: Standard OAuth 2.0, No Surprises
One of the most practical aspects of this integration is its authentication model. Rather than introducing a proprietary credential mechanism, Oracle has implemented standard Google Cloud OAuth 2.0 authentication for Select AI profiles targeting Gemini Enterprise Agent Platform.
For teams already managing Google Cloud credentials as part of their Oracle Database@Google Cloud deployments, this means the integration slots naturally into existing identity and access management workflows. There’s no new credential type to learn, no additional secrets management layer to deploy — just standard OAuth 2.0 tokens scoped to the appropriate Google Cloud project.
Keeping Pace with Google’s Platform Evolution
This update also reflects Oracle’s commitment to staying current with its partners’ evolving platform strategies. In April 2026, Google officially rebranded Vertex AI as the Gemini Enterprise Agent Platform, consolidating its AI and agent-building capabilities under the Gemini brand umbrella. Oracle’s decision to support this rebranded platform promptly signals that the Oracle-Google multicloud partnership is actively maintained, not merely a static checkbox.
For organizations that had previously configured Select AI profiles using earlier Google AI provider options, it’s worth reviewing your configurations to take advantage of the latest Gemini models and capabilities now accessible through the updated platform integration.
Part of a Broader October 2026 Release
The Gemini Enterprise Agent Platform support didn’t arrive in isolation. It was released alongside several other notable Autonomous AI Database enhancements in the October 2026 update cycle:
- Iceberg External Table Partition Pruning: A performance-focused improvement for lakehouse query workloads. Queries against Apache Iceberg external tables can now leverage partition pruning, significantly reducing the amount of data scanned and improving query response times for analytical workloads that span the data lakehouse.
- Oracle APEX 26.1: The latest release of Oracle’s low-code application development platform, bringing new features and improvements for developers building data-driven applications on Autonomous AI Database.
Taken together, these updates paint a picture of Oracle’s Autonomous AI Database as a platform that is maturing rapidly across multiple fronts — AI integration, lakehouse performance, and application development tooling.
Practical Takeaway for Oracle Professionals
If your organization runs Oracle Database@Google Cloud and you’ve been building or evaluating in-database AI workflows, this integration removes a significant barrier. You can now point Select AI at Google’s most capable Gemini models using straightforward profile configuration and standard OAuth 2.0 credentials — no external API gateway, no middleware layer, no custom code.
The action item is clear: review your current Select AI profiles, evaluate whether Gemini Enterprise Agent Platform models offer improvements over your current provider for NL2SQL accuracy, embedding quality, or conversational capabilities, and test the integration in a non-production environment. For teams already invested in the Google Cloud ecosystem, this is the most frictionless path to bringing enterprise-grade generative AI directly into your Oracle SQL workflows.
Oracle’s in-database AI story continues to evolve, and with each new provider integration, the case for keeping AI processing close to the data — rather than shuttling data to external AI services — grows stronger.