Select AI
Oracle Autonomous AI Database Select AI is Oracle's built-in feature that lets you interact with your data using natural language instead of writing SQL, and then uses AI to generate, run, and even explain the query for you.
Extend Select AI Translate for Google, AWS, and Azure Translation Services
Select AI extends the set of translation services supported by the translate action to include services from Google, AWS, and Microsoft Azure. This expands the existing support for the OCI translation service (OCI Language).
This feature enables developers to easily incorporate language translation capabilities into applications for use cases such as translating global news articles, social media content, product documentation, and other content. Translation can be pipelined with other Select AI functionality to handle LLM limitations or customize output for individual users or regions.
Select AI Capabilities
Through the DBMS_CLOUD_AI package, Select AI serves as an AI gateway enabling users, developers, and applications to interact with Oracle AI Database using natural language supporting multiple generative AI use cases. Select AI provides functionality supporting natural language to SQL, including automatically generating, running, and explaining SQL; supporting automated vector index creation and the retrieval augmented generation (RAG) workflow; and through the DBMS_CLOUD_AI_AGENT package, providing a framework for building, managing, and running AI agents in the database. It also supports synthetic data generation and text summarization and translation. Select AI provides PL/SQL and Python APIs.
Together with the broader foundational Cloud Developer packages, they enable you to integrate AI capabilities directly from the database and support end-to-end workflows that deliver greater data insight with AI.
Select AI enables business users and developers to use a wide range of AI providers and models supporting many AI-enabled use cases and application scenarios. Select AI helps to simplify the adoption of AI in enterprises and organizations across both on-premises and cloud systems. Users get insights faster using natural language without waiting on manual SQL query generation or workflows implementing retrieval augmented generation. This can reduce IT bottlenecks and accelerate AI-enabled application development. Using Select AI, enterprises benefit from database security mechanisms and protocols.
Select AI Support for NL to Graph Queries
Select AI extends table-based natural language to SQL generation to include native database property graphs. This enables answering questions like "Search for persons or persons they know where both people have linked the same post" or "Find the names of people they know (via 3 to 7 other people)."
Select AI support for native database property graphs lets developers expose graph-based insights using familiar database and natural-language interfaces, reducing custom integration and query-development effort. Business users can explore complex relationships, dependencies, and patterns without learning graph query languages—accelerating analysis and improving decision-making while retaining database security and governance.
Select AI Supports Specifying Custom LLM via Private Endpoint in Customer-Hosted Environment
This feature enables you to run your own LLM on a VM controlled by your organization and configure its private endpoint in an AI Profile for use with Select AI.
You gain greater control over model choice, customization, data residency, and security while minimizing data movement outside your private environment. Developers can integrate organization-specific models through a familiar AI Profile configuration, and business users benefit from AI experiences tailored to their domain and policies.