Oracle: AI applications as simple as possible with the AI Data Platform
Oracle introduces a central data platform and marketplace for AI agents. They are intended to simplify the development and operation of AI cloud applications.
(Image: dpa, Ralf Hirschberger)
- Arne Bauer
- Jens Söldner
Oracle unveiled several innovations for businesses at its flagship event, Oracle AI World, in Las Vegas. With the new AI Data Platform and an AI Agent Marketplace for Fusion Cloud Applications, the company aims to simplify and standardize the use of AI in the corporate environment.
According to the manufacturer, the Oracle AI Data Platform is designed for building and operating AI applications. It combines automated data ingestion, semantic optimization, and vector indexing with integrated generative AI tools. This should enable companies to transform raw data into actionable insights faster and integrate their AI agents into existing workflows.
Several Oracle components are used, including the cloud infrastructure (OCI), the autonomous AI database, and the generative AI service. The platform supports open Lakehouse formats such as Delta Lake and Iceberg and offers zero-ETL and zero-copy access to operational data from finance, HR, or supply chain systems. An IT service catalog is also intended to enable unified governance across all data and AI assets. The so-called Agent Hub serves as the central control point; it evaluates requests, forwards them to the appropriate agents, and bundles the results.
Videos by heise
Expansion of Fusion Cloud Applications
In addition, Oracle is expanding its Fusion Cloud Applications with pre-built agents, including those for financial planning, invoice processing, and HR talent management. If the agents are not sufficient for the required scenario, the manufacturer is introducing another source with the AI Agent Marketplace. Partner companies such as Accenture or Infosys, as well as software providers like Box or Stripe, offer specialized AI agents there as verified and ready-to-use templates. All agents can work directly within existing workflows, analyze data in real-time, provide recommendations, and automate recurring tasks.
Finally, the AI Agent Studio has also been expanded. It now supports multiple large language models, including OpenAI, Anthropic, Cohere, Google, Meta, and xAI. New features are intended to cover the entire lifecycle of agents, from creation and testing to observation and operation. This includes monitoring dashboards, prompt management, multimodal RAG, and a credential store for saving API keys and tokens.
More information about the announcements can be found here:
(vbr)
