Use this matrix to move from a DAMA domain demand to an action with Oracle: identify the owner, select the capacity and define the evidence of the result. The same decision can combine processes, people and several technical controls. The official links allow to deepen the configuration and use of each function.
Conceptual and Technology Architecture
The conceptual plane begins with a data product: customers and contracts suitable for billing, indicators and authorized care. Define owner, aims, schemes, quality standards, frequency, consumers and procedure of change. The sources retain their responsibilities. The authorisation of the analytical product does not replace the policies of operational systems.
At the technological level, Oracle AI Data Platform organizes assets, execution, permissions and engineering evidence. Oracle Autonomous AI Lakehouse provides SQL processing, analytics and integration with open formats. Connectors, credentials, networks and formats determine each relationship; The arrows represent roads that have to be configured and validated. Knowledge Bases, Agent Flows and AI Compute Oracle AI Data Platform require an instance of Oracle Autonomous AI Lakehouse 26ai or more, as AI function requirements. The ingestion and processing of structured data can be used without enabling these functions.
Oracle Cloud Infrastructure GoldenGate is added as necessary to capture changes. Data Studio from Oracle Autonomous AI Database and Oracle Data Transforms allow you to explore and prepare data in your area. Oracle Analytics Cloud uses datasets with revised semantics. Oracle Data Safe applies to supported databases Oracle Fusion Cloud Enterprise Data Management is a SaaS application with its own requirements. Oracle Cloud Infrastructure Identity and Access Management, keys, storage and observability hold concrete controls, not a global authorization that spreads to everything.
Decisions Matrix and Evidence
Read every row as an app: what business needs, who decides and how it's implemented. For example, an approved goal becomes permissions about a product, access tests and audit records.
| DAMA area | Requirement | Responsible | Application with Oracle |
|---|---|---|---|
| Data Government | Assign Access to Assets and Review Decisions | Owner of data Administrator | Oracle AI Data Platform: roles and permissions objects the approval of an aim remains organisational Official documentation: |
| Data Government | Check who operated an asset | Auditor | Oracle AI Data Platform: supported operations audit records Official documentation: |
| Data architecture | Separate consumption production SQL | Architect | Oracle AI Data Platform + Oracle Autonomous AI Lakehouse; explicit connections and prerequisites Official documentation: |
| Modeling Data and Design | Define Analytical Institutions and Semantics | Modeler Business | Tables and views at Oracle Autonomous AI Database Analytical views for measurements and dimensions Official documentation: |
| Modeling Data and Design | Publish terms and measures shared | Modeler owner from data | Oracle Analytics Cloud Semantic Modeler: To relate tables, to define cardinality and aggregations and to publish the consumption model Official documentation: |
| Storage from Data and Operations | To retain objects according to a policy | Documentary Officer operations | Oracle Cloud Infrastructure Object Storage: retention rules Distinguish them from versions and lifecycle Official documentation: |
| Storage from Data and Operations | Detect detour and assign expenses | Operations FinOps | Metrics, Oracle Cloud Infrastructure's Cost Analysis and budgets budget alerts do not cut consumption Official documentation: |
| Security of Data | Separate identities and cryptographic material | Security platform | Oracle Cloud Infrastructure Identity and Access Management and Oracle Cloud Infrastructure Key Management Service permissions from platform and data have to be reviewed separately Official documentation: |
| Security of Data | Discover pII and prepare test copies | Security DBA | Oracle Data Safe: Uncover sensitive attributes, configure subsetting and masking in a registered copy and review relationships and reports Official documentation: |
| Security of Data | Classifying and protecting data with human review | Steward Security | Implement a flow onto Oracle AI Data Platform: classify attributes, approve treatments, generate protected views or masked copies and check permissions Official documentation: |
| Data integration and interoperability | Capture changes with lower latency | Data engineering | Oracle Cloud Infrastructure GoldenGate: Oracle AI Data Platform connection when source, destination and pattern are supported Official documentation: |
| Data integration and interoperability | To prepare data and coordinate units | Data engineering | Oracle AI Data Platform and Oracle Data Transforms jobs in their fields AI's notations and assistance with its own configuration Official documentation: |
| Documents Management and Content | Reget authorized knowledge for AI | Documentary owner | Knowledge Oracle AI Data Platform bases: ingestion and recovery of supported content preserve valid policy Official documentation: |
| Data Master and Reference | Approve changes and solve duplicates | Data steward owner | Oracle Fusion Cloud Enterprise Data Management: match / merge and survival rules evaluate domain, scale and SaaS requirements Official documentation: |
| Data Warehousing and Business Intelligence | Publish consistent indicators | Responsible for BI | Oracle Autonomous AI Lakehouse + Oracle Analytics Cloud; with documented Oracle AI Data Platform Official documentation: |
| Data Warehousing and Business Intelligence | To assist with preparations and analytics | Analyst Oracle Analytics Cloud manager | To enable AI Assistant to generate columns and expressions in data flows; using MCP to discover sources, check metadata and run ANSI SQL, revising filters and results Official documentation: |
| Metadata management | Find and manage Oracle AI Data Platform assets | Steward engineering | Master Catalog from Oracle AI Data Platform Official documentation: |
| Metadata management | Analyzing Units and Changes | Engineering owner | Oracle AI Data Platform lineage of entities and columns and impact analysis of supported artifacts last registration by process Official documentation: |
| Metadata management | Consult sources with assembled catalogs | Architect DBA | Federated catalog in Oracle Autonomous AI Database; with DBMS_CATALOG Official documentation: |
| Metadata management | share Iceberg transactional metadata | The lakehouse platform | Oracle AI Data Catalog: REST service administered within Oracle Autonomous AI Database, other than Oracle AI Data Platform's Master Catalog Official documentation: |
| Metadata management | Enriching descriptions and revising impact | Steward | Data Studio Oracle Autonomous AI Database: exploration and supported lineage Oracle Data Transforms: Business and AI Support Official documentation: |
| Quality of Data | Accept, Reject and Rewrite by Business Rules | Owner engineering | Oracle AI Data Platform: run SQL checks at notebooks, coordinate validation and publication with workflows and register Rejections Official documentation: |
Three Catalogs and a Common Responsibility
Master Catalog from Oracle AI Data Platform organizes Oracle AI Data Platform's assets. The Federation catalogue of Oracle Autonomous AI Database organizes access to sources by means of assembly. Oracle AI Data Catalog implements Iceberg's REST interface for tables metadata and transactions. Access to storage is an additional check. None of these names prove that the three of them are replying to each other.
At Horizonte, each connection records origin, effective identity, exposed objects, permissions, update and owner. If a corporate term with tools needs to be shared, its identifier, version and publication process will be defined. does not be decided by enabling a catalogue.
The lineage helps to find consumers of a change but does not replace a historic registration of approvals. A suggestion from AI can enrich a description. steward values its meaning before its publication. The solution should show the complete string: decision → control → execution → evidence → review.