Oracle adds semantic layer to Autonomous AI Lakehouse

Oracle has added new features to its Autonomous AI Lakehouse, extending support for Apache Iceberg data and adding a business semantic layer.

The changes are intended to give users and AI agents more business context when working with enterprise data.

The latest additions are designed to let organisations work with real-time operational data alongside data stored in Iceberg tables, while applying common governance rules across different systems. Customers can do this without moving data into a separate repository.

The new functions cover data discovery, analytics, governance and business planning. The offering is available across Oracle Cloud Infrastructure, Amazon Web Services, Microsoft Azure, Google Cloud, Oracle Exadata Cloud@Customer and OCI Dedicated Region.

Business context

A central part of the update is a Business Semantic Layer, which helps users and AI agents query and interpret enterprise data with consistent definitions. The feature uses catalog annotations, imported SQL annotations and AI-suggested analytic views to turn raw data into business-aware information.

Those reusable analytic views can include dimensions, measures, metrics, hierarchies, calculations and key performance indicators. The aim is to improve the accuracy of natural-language querying, AI retrieval and analytics.

Oracle is also adding an AI Data Catalog, which it says gives customers a broader view of enterprise data by allowing them to mount external catalogs, Oracle and non-Oracle databases, Iceberg tables, applications and object store data without duplicating it.

According to Oracle, the catalog can query these sources live through the Oracle engine. Agentic functions in the catalog can enrich metadata with business semantics, lineage, workload patterns, relationships and retrieval signals.

Oracle said the AI Data Catalog and Business Semantic Layer will be available within the next 12 months.

Iceberg support

Oracle has also expanded support for Apache Iceberg, an open table format used for large analytic data sets. Customers can create, populate, update and delete Iceberg tables using SQL, giving them a common analytics model across Oracle and Iceberg data.

Organisations can use those tables for AI-driven analytics and agent workflows, and create reusable data products with time-travel functions for audit and recovery. Oracle added that native Iceberg support has been enhanced to improve performance and broaden functionality for users accessing tables through external catalogs or the Oracle AI Data Catalog.

Data Studio is another part of the release. The workspace lets teams discover, prepare, analyse and manage lakehouse data without first provisioning a database. Users can connect to more than 100 data sources and work with data in natural language, SQL or Python.

Controls and planning

Oracle also introduced what it calls centralised security and controls. A single control plane in the AI Data Catalog allows customers to apply role-based and object-level access controls across Oracle, non-Oracle and Iceberg data in both cloud and on-premises environments.

Another addition is a Business Simulation Engine, built using Oracle Essbase inside the lakehouse. Oracle said the engine supports multidimensional simulations and what-if analysis, including write-back, dynamic hierarchy rollups, advanced calculations and allocations.

The tool is aimed at planning and forecasting. Customers can ask questions in natural language, generate simulation reports and analyse data in Excel using information stored in Oracle databases or files in object storage.

Oracle also updated its Data Lake Accelerator. Benchmark tests on 100 TB TPC-DS data sets showed up to three times faster query execution, while the maximum cluster size for the accelerator has also increased threefold.

Customer use

Liberty Energy, an energy services company in North America serving oil and natural gas producers, is using Oracle Autonomous AI Lakehouse across a multicloud environment, according to Oracle. The deployment brings together financial, operational, supply chain and workforce data.

"Giving AI access to enterprise data is not enough. To make informed decisions and take trusted action, AI agents need to understand the business context behind that data and operate within an organization's rules and policies," said Çetin Özbütün, Executive Vice President, Autonomous AI Database Technologies, Oracle.

"By bringing business semantics and multidimensional modeling together with open data access and unified data governance, the latest updates to Autonomous AI Lakehouse help organizations give AI agents the context and controls to interpret enterprise data, apply business rules, and support trusted actions."

Liberty Energy outlined how it is using the platform across multiple systems.

"Oracle Autonomous AI Lakehouse gives decision-makers real-time visibility into field revenue, inventory usage, costs, and profitability," said Rhett Rowan, Vice President of IT, Liberty Energy.

"It seamlessly brings together Apache Iceberg tables with data from Amazon S3, Amazon databases, Microsoft SQL Server on Azure, MongoDB, and PostgreSQL, making it easy to add new data sources to power AI, analytics, and business simulations."

Industry analyst Dave Vellante said the latest step reflects a broader shift in enterprise AI.

"We believe the next phase of enterprise AI will depend on connecting data access with the business context needed to make useful decisions," said Dave Vellante, Co-Founder and Chief Analyst, theCUBE.

"Oracle's Autonomous AI Lakehouse takes a significant step in that direction by bringing together governed access to Oracle and non-Oracle data, shared business definitions, AI-assisted analytics and multidimensional what-if planning. This combination can help enterprises ground agent workflows in the metrics and relationships that matter to their business and evaluate options using a more consistent analytical foundation. Realizing that value requires accurate business definitions and clear controls over execution. Oracle is strengthening the lakehouse as a practical foundation for planning, informed decisions and agentic workflows."