Business semantic layer for AIDP catalog that is complimentary to OAC
Business semantics (metrics, KPIs, business entities, hierarchies) currently live primarily inside downstream analytics tools such as Oracle Analytics Cloud, or are redefined separately in each consuming platform leading to a fragmentation of definitions of key metrics, no cross-platform single pane of glass for business meaning and AI agents working off schemas instead of business views. Governance and catalogs are centralised in AIDP but business semantics isn't.
Introduce a business semantic layer in Oracle AI Data Platform (AIDP) that sits on top of the AIDP master catalog and external catalogs, and is consumable by Oracle Analytics Cloud (OAC), AI agents and other analytics platforms. This is not a replacement for the OAC semantic model, but an extension of it and a shared, enterprise-wide semantic “single pane of glass” across all data products: structured, semi-structured and unstructured.
Key benefits
- Single pane of glass across all enterprise data
- One place to define business entities and metrics across structured warehouse data, semi-structured lakehouse data and unstructured AI-ready assets.
- OAC and other analytics platforms consume semantics from AIDP instead of recreating them.
- Consistency of metrics and definitions
- “Revenue” or “Churn” means the same thing in OAC dashboards, FDI/FAIDP, AI agents, and other Analytics platforms because they all point back to the AIDP semantic layer.
- Better AI agent behaviour and governance
- Agents reason over governed business concepts rather than arbitrary table and column names.
- Natural language questions can be resolved to semantic entities and metrics that are already secured and documented.
- Reduced duplication, lower maintenance cost
- Less repeated effort defining metrics and hierarchies in every downstream consumer tool.
- Changes to core metrics are made once in AIDP and then flow through to all consumers.
- Stronger alignment between AIDP and OAC
- AIDP owns the catalog, governance and cross-source semantics.
- OAC continues to own rich analytics and visualisation semantics.
- The two layers work together rather than overlap, reinforcing Oracle’s “single platform” story.
AIDP is already framed as the governed, single pane of glass over the enterprise data ecosystem, across multiple storage and compute engines. Extending it with a business semantic layer completes the loop with data products no longer being viewed as just technical artefacts but governed business concepts.