Oracle boosts AI Lakehouse and Iceberg speed

Oracle (ORCL) has rolled out new Autonomous AI Lakehouse capabilities designed to combine business context for AI agents with expanded support for Apache Iceberg data. The updates are intended to improve how organizations discover, analyze, and govern data across both Oracle and non-Oracle sources while limiting the need for costly data movement.

A central addition is a Business Semantic Layer that provides shared definitions of business metrics for both users and AI agents. This layer is aimed at grounding AI-driven workflows in consistent business logic so that analytics, reporting, and automated decisions rely on the same core definitions.

Oracle reports that the maximum Data Lake Accelerator cluster size has increased threefold. In benchmark tests using 100 TB TPC-DS data sets, Iceberg queries achieved up to threefold faster execution, highlighting a focus on scaling performance for large, analytics-heavy workloads.

Expanded Apache Iceberg table management is part of the new release, supporting broader use of Iceberg as a format for analytics and AI. These Iceberg improvements sit alongside the broader Lakehouse enhancements, positioning the platform to handle larger data volumes with improved query speeds.

Oracle states that the Business Semantic Layer and AI Data Catalog capabilities are expected to be available within the next 12 months. The company also notes that release timing is not guaranteed and may change, framing the timeframe as a target rather than a firm commitment.

The AI Data Catalog is intended to support richer data discovery and governance across an enterprise data estate. Together with the semantic layer, it is designed to help align data assets, access policies, and AI agents with consistent business meanings and rules.

The Autonomous AI Lakehouse update includes a Business Simulation Engine that integrates Oracle Essbase to support forecasting and what-if analysis. This capability targets multidimensional simulations, allowing organizations to model potential business outcomes using their existing data.

An AI-assisted Data Studio workspace is also introduced to streamline how users work with data within the Lakehouse. Across these enhancements, Oracle positions the Lakehouse as a unified environment where AI agents, semantic definitions, performance-optimized Iceberg data, and simulation tools operate on a governed, enterprise-scale data foundation.