Oracle already has over 1,000 AI agents running in Fusion Cloud applications. Now comes a set of 22 Agentic Applications, tools that go beyond regular agents by operating much more autonomously. Oracle hopes this will mark the transition to a “system of outcomes.” How realistic is that?
Whereas the company previously focused on copilots and AI assistants as add-ons to existing software, the Fusion Agentic Applications serve as a separate layer that “actively pursues results,” Oracle explains. Agents have access to business data, workflows, rules, and context. In other words: everything needed to complete processes without human intervention for routine tasks. How successful they are in this will depend just as much on the quality of user input as on the models and structures behind the Agentic Applications.
“With Fusion Agentic Applications, we’re moving enterprise software beyond passive systems of record and offering our customers applications that can reason, decide, and act within the context of defined business objectives,” said Steve Miranda, executive vice president of Applications Development at Oracle.
22 applications for finance, HR, and supply chain
There are currently 22 Fusion Agentic Applications available, targeting a wide range of business functions. For example, the Workforce Operations Agentic Application helps HR teams reduce payroll issues and accelerate scheduling approvals. The Collectors Workspace Agentic Application supports finance teams in collecting payments faster and reducing days sales outstanding. For supply chain teams, there is the Design-to-Source Workspace Agentic Application, which brings engineering, supplier, and procurement decisions together into a single coordinated process.
Furthermore, organizations can also get started on their own using the AI Agent Studio. More enhancements are available here. Oracle introduced this tooling a year ago as a platform for building AI agents within Fusion Applications, and expanded it in October with an AI Agent Marketplace. A further expansion followed in February with 29 pre-built AI agents for supply chain and customer experience. The Fusion Agentic Applications take it a step further: instead of standalone agents, the focus is now on coordinated applications with a shared, continuous context that retains the intent, history, and decisions from previous steps.
The new Agentic Applications Builder enables organizations to build results-oriented agentic applications. This is possible through a natural language-based environment, without traditional application development. Additionally, new capabilities have been added for workflow orchestration, context memory, content intelligence, and LLM multimodal functionality. This allows agents to process images, audio, and video as well. A new Agent ROI dashboard measures time savings and productivity gains per agent.