nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Oracle Communities

Oracle AI and Analytics

Child Item

AI Agent - Subject Area Usage

Saroj Kumar Behera

Hi,

Can we create OAC - AI Agent based on Subject area data.

Regards

Saroj

Find more posts tagged with

Agent

OAC

Accepted answers

RVohra

@Saroj Kumar Behera ,

We can definitely create AI Agents based on the Subject Area but need to consider following:

AI agents cannot directly query the database. They rely on the semantic model / subject area.
Subject areas must have:
1. good metadata
2. friendly column names
3. defined hierarchies
Security still applies:
- row-level security
- role-based access

Hope it helps!

All comments

Rajaraman Loganathan

Yes, OAC AI Agent can be created on Subject Area but to enable creating AI Agent based on specific Subject Area Indexing needs to be defined on the said subject area using Search.Index option.

Mohammed Rafi

@Rajaraman Loganathan - If I am not wrong there is no separate indexing step required to make a Subject Area accessible for Natural Language Query (NLQ) or AI Assistant.

Rajaraman Loganathan

@Mohammed Rafi Could see only the Indexed subject areas are listed to choose dataset while creating AI Agent

Brendan Tierney

Recent updates to the platform have specifically expanded the AI Assistant and AI Agent frameworks to support enterprise subject areas alongside standard datasets.
Key Features for Subject Area-Based AI Agents
Indexing and Synonyms: As of March 2025, search indexing includes support for subject areas. This allows administrators to define synonyms for columns within a subject area, making it easier for the AI Agent to understand natural language business queries (e.g., mapping "Earnings" to the "Converted Revenue" column).Author Control: Content authors can explicitly specify which data sources—including specific subject areas—users are permitted to interact with using natural language.
Domain Specialization: The January 2026 update introduced domain-specialized AI agents. These agents leverage Retrieval-Augmented Generation (RAG) and custom instructions to allow consumers to interact with curated enterprise data models (subject areas) using their own business vocabulary.

RVohra

@Saroj Kumar Behera ,

We can definitely create AI Agents based on the Subject Area but need to consider following:

AI agents cannot directly query the database. They rely on the semantic model / subject area.
Subject areas must have:
1. good metadata
2. friendly column names
3. defined hierarchies
Security still applies:
- row-level security
- role-based access

Hope it helps!

Upcoming Event

AI Data Platform - Creating High Code Agents and Using Developer Tools in AI Data Platform
Explore how to build agents using code in AI Data Platform. We'll create an agent that examines and analyzes the contents of a Github repository and reports its findings. The session demonstrates how to use different developer tools to create agents including the AIDP extension for VS Code and Codex. Using these tools we…

Top Leaders

Mallikarjuna Kuppauru-Oracle 80