Select AI provides developers, analysts, and application teams with a natural language layer on top of enterprise data. This allows you to generate SQL, chat with your data, tailor responses with authoritative content, create synthetic data, and build agent workflows directly within your database. Oracle’s current Select AI feature matrix shows that this feature set is available for both Oracle Autonomous AI Database and Oracle AI Database, with some advanced features concentrated in the new 26ai release.
Equally important, Select AI is included in Oracle AI Database and Oracle Autonomous AI Database. Deliver the AI model you want to use by choosing from a wide range of supported providers and models that suit your requirements. Oracle documentation and product documentation emphasize this flexibility and list support for multiple providers, including OCI Generative AI, OpenAI, Azure OpenAI, Cohere, Google, Anthropic, Hugging Face, Amazon, OpenAI compatible providers, and private LLM deployments, depending on the release.
Across both Oracle Autonomous AI Database and Oracle AI Database, Select AI provides a consistent core experience in both 26ai and 19c, including chat, natural language to SQL generation (NL2SQL), synthetic data generation, AI agents, summarization, translation, and more.
The 26ai release is built on a common foundation with the addition of NL2SQL feedback, automatic object selection, and search enhancement generation (RAG). These features rely on Oracle AI Vector Search, which is not available in 19c.
For Oracle AI Database, version number 23.26.1 corresponds to the 26ai release.
This means organizations will standardize on: 26ai While getting an enhanced Select AI experience. 19c We provide a powerful baseline that includes chat, NL2SQL, synthetic data generation, AI agents, summarization, and translation.
NL2SQL
Natural language to SQL is a core feature of Select AI that allows users to describe desired data results in natural language and generate, execute, narrate, and explain the resulting SQL.[AI はスキーマ メタデータを使用して、AI プロファイルで指定された LLM によって使用される拡張プロンプトを生成する]Select. For developers, NL2SQL generation speeds query creation, and for business users, it enables access to enterprise data without requiring deep SQL expertise.
feedback
In 26ai, Select AI adds feedback NL2SQL allows users to provide feedback on the generated results via the SQL command line or PL/SQL procedures. The goal is simple. The goal is to improve query generation accuracy over time by learning from user modifications and settings. This makes Select AI more adaptable for use in real-world enterprises, where schema complexity and business language often evolve simultaneously.
automatic object selection
Additionally, with 26ai, Select AI for NL2SQL can automatically select table metadata relevant to a prompt and send only the required metadata to LLM. This uses automatically created vector indices for objects specified in the AI profile. Based on user prompts, semantic similarity search helps identify metadata relevant to a query. This simplifies prompt creation and improves accuracy by narrowing the context to the most relevant objects.
rug
Select AI Search Enhancement Generation combines LLM inference with trusted enterprise content. Oracle automates the RAG flow from generating the embedding to retrieving semantically similar content from the vector store and augmenting the prompt with that data. The result is more relevant and grounded answers with less risk of hallucinations, especially for personal and domain-specific information.
SDGs
Synthetic data generation allows teams to create schema-compliant data for development, testing, UX validation, and machine learning projects without exposing sensitive production data. Use cases include creating metadata clones, activating new projects when real data is not available, supporting AI/ML use cases with limited data access, and more.
AI agent
The Select AI Agent framework extends the platform from instant response to action-oriented workflows. It is an autonomous agent framework for building and managing agents within a database that allows agents to reason about requests, invoke built-in or custom tools, reflect on results, and maintain short- and long-term memory across conversations. In practice, this means that developers can create agent workflows using SQL and PL/SQL while keeping logic, governance, and data tight.
summary
Summaries are one of the features that are immediately useful on a daily basis. With the specified LLM, Select AI can return text responses, summarize query results in prose, and enable conversational follow-up. This makes it easy to convert database output into readable business language for reports, dashboards, and embedded applications.
translation
Translation extends the value of natural language interaction by helping teams generate responses in multiple languages and adapting content to users around the world. Users can leverage translation services from OCI, GCP, AWS, and Azure.
chat
Chat provides the basic interaction model for Select AI, allowing users to submit prompts ranging from simple questions to complex requests. These prompts are passed directly to LLM unchanged, so the user’s original intent is preserved.
Conversations, which apply to all Select AI features, build on that foundation by supporting multi-turn interactions between users and LLMs, maintaining context across prompts, and enabling more natural follow-up interactions.
The leap to 26ai is significant because it transforms Select AI from a powerful natural language interface to a more complete AI application layer.
- Feedback improves NL2SQL’s accuracy over time
- Automatic object selection reduces manual schema targeting
- RAG grounds responses within trusted enterprise content
Combined with the AI agent, 26ai makes Select AI a particularly attractive release for enterprise-grade AI applications and workflows.
The message is simple: Select AI is available for both Oracle Autonomous AI Database and Oracle AI Database, with 26ai offering the broadest feature set. And because AI models are delivered separately, you can combine Oracle’s in-database AI framework with the provider and model strategy that best fits your performance, governance, and cost requirements.
For more information…
