
AI Data Platform AI Registry
Oracle AI Data Platform gives organizations one governed place to register, discover, and manage enterprise AI capabilities. Administrators use AI Registry to register agents, MCP servers, tools, and skills in a shared inventory, along with information such as ownership, dependencies, access permissions, and required checks. Teams can then find and reuse existing capabilities, monitor usage and costs, and manage their lifecycle with the information and controls needed to make them available across applications and workflows.
Why Oracle AI Data Platform AI Registry?
-
Manage AI growth with visibility and control
Bring agents, MCP servers, tools, and skills into a shared inventory. Track ownership, dependencies, and access permissions to manage AI capabilities as adoption grows.
-
Find and reuse existing AI capabilities
Use descriptions and metadata to find existing capabilities, understand their purpose, and check who can access them. Reduce duplicate development and make informed choices about what to reuse.
-
Connect an open ecosystem to a governed experience
Register fully A2A-compliant agents from Oracle and non-Oracle applications and make approved capabilities discoverable and available according to access rules.

How AI Registry works
AI Registry provides a shared foundation for discovering, governing, and reusing enterprise AI capabilities. Teams register capabilities, describe what they do, and manage their availability and access through supported applications and assistant interfaces.
Explore AI Registry in AI Data Platform
AI Registry gives teams a central place to register, organize, and govern enterprise AI capabilities. Administrators manage ownership and access, developers discover capabilities they can reuse, and users access approved agents through supported experiences that fit their work.
Bring agents into AI Registry
Register agents from across the enterprise
Add any fully A2A-compliant agent, whether it was built in AI Data Platform, an Oracle application, or another agent platform.
Register reusable tools and skills
Register approved MCP servers, tools, and reusable skills so capabilities can be shared across agents rather than rebuilt for every project.
Describe capabilities for discovery and routing
Capture an A2A agent card and quality metadata, such as name, description, owner, domain, authentication, and supported capabilities, so users and routing services can identify the right agent.
Organize capabilities around business needs
Group related capabilities into Topics
Organize agents and supporting capabilities around a business area or use case, helping administrators structure what is available.
Provide context for discovery
Topics provide additional context for identifying capabilities relevant to a user’s request.
Bring relevant capabilities together
A finance Topic, for example, can bring together invoice and budget agents, forecasting tools, and relevant finance data.
Govern the enterprise agent estate
Control availability with roles
Apply role-based access controls to registered agents, MCP servers, and tools so only authorized users can discover and use each capability.
Establish ownership and accountability
Use consistent metadata to identify who registered an agent, what business purpose it serves, and which systems or capabilities it depends on.
Manage capabilities throughout their lifecycle
Track versions and required policy checks and support testing and review as registered capabilities evolve.
Monitor behavior, usage, and cost
Gain visibility into how registered capabilities behave, how often they are used, and their associated costs to inform ongoing management.
Access agents and tools through supported experiences
Choose an access method that fits the work
Use AI Registry to access your agents through tools such as ChatGPT, Codex, Claude, or Oracle AI Data Platform Agent Hub.
Use registry-informed discovery and routing
Provide descriptions, capability metadata, and access information that supported experiences can use to identify relevant agents and tools for a request.
Work conversationally through Agent Hub
Use Agent Hub to ask questions, request actions, and interact with approved agents or MCP tools through a consistent conversational experience.
Continue work with context and oversight
Preserve sessions
Let users return to ongoing work with prior interaction context available instead of restarting every agent conversation.
Retain agent-generated outputs
Keep the outputs produced during an interaction—such as reports, analyses, or other response artifacts—without turning Agent Hub into the store for the underlying enterprise data.
Support enterprise oversight
Use enterprise identity, logging, and tracing to understand activity. Required human approvals remain part of the underlying agent or business workflow.
추천 AI Data Platform 블로그
에이전트 vs. 워크플로: ROI는 실제로 어디에서 발생할까요?
대부분의 기업이 AI 에이전트 도입에 실패하는 이유는 기술 때문이 아니라, 단순한 워크플로로 더 잘 처리할 수 있는 업무에 에이전트를 사용하고 있기 때문입니다. 이 블로그에서는 워크플로와 자율형 에이전트의 주요 차이점과, 올바른 접근 방식을 선택하는 것이 ROI, 확장성, 거버넌스 및 비용에 어떤 영향을 미치는지 설명합니다. 또한 에이전틱 AI가 실제로 투자할 가치가 있는지 판단할 수 있는 실용적인 프레임워크를 제공합니다.
-
2026년 9월 18일
에이전트, 툴, 스킬: 엔터프라이즈 AI를 위한 더 효율적인 접근 방식
-
2026년 8월 12일
귀사의 데이터는 이미 해답을 알고 있습니다. 이제 그 해답을 바탕으로 행동할 수 있습니다.
-
2026년 6월 30일
Oracle, ISG Research AI 및 데이터 플랫폼 부문 전반에서 선도적 입지 확보
