Oracle Fusion unveils AI agents to boost finance team efficiency

Oracle has unveiled new AI agents within its Oracle Fusion Cloud Applications, targeting enhanced operational efficiency and business insight for finance teams. The company says the move will help businesses automate complex financial workflows while maintaining the highest standards of security and governance.

Oracle's Enterprise Resource Planning (ERP) and Oracle Fusion Cloud Enterprise Performance Management (EPM) suites now span more than 11,000 global customers, serving large enterprises as well as small and medium-sized businesses. By embedding AI within both transaction and planning processes, Oracle aims to bridge traditional accounting workflows with predictive, data-driven insight.

"Oracle is ushering in a new era of agent-driven finance, where AI assistants turn fragmented, complex staff-heavy processes into proactive, continuous operations," said Rondy Ng, Executive Vice President of Applications Development at Oracle. He added that the agents "free teams to focus on judgment and strategic outcomes."

Running on Oracle Cloud Infrastructure, the agents are prebuilt with advanced security and embedded natively within existing workflows, helping users make faster, more informed decisions. The Planning Agent, for example, can run "what-if" simulations, guiding finance leaders through real-time scenarios to understand the profitability and operational implications of strategic decisions.

Hari Sankar, Group Vice President of Product Management at Oracle, said the embedded approach marks a significant step forward in how finance teams can safely adopt AI.

"What Oracle has done is to take the time to build AI capabilities fundamentally into the application layer," he explained. "Embedding AI capabilities into the platform layer which allows us to obey all of the existing security, access controls, authorisations and governance policies."

He said this design helps address one of the finance sector's biggest concerns: the handling of sensitive data. "We maintain clean separation between your data and somebody else's data...we ensure the usage is very customer-specific."

Trust, Sankar noted, is a process that develops gradually. Oracle's engineers are focused on helping users understand not just what AI predicts, but why it makes a specific decision.

The AI agents are being positioned as tools to help users operate with greater speed and accuracy while maintaining compliance and governance standards. Each of the AI agents addresses different areas in the finance function:

The Payables Agent is designed to automate accounts payable processes. It can process invoices from various sources, including email, portals, EDI/e-invoicing, and PDFs. The agent can extract and normalise data, match it to purchase orders and receipts, create accounting entries, and apply relevant tax, policy, and fraud checks before routing for approval and payment. Oracle states that this will reduce manual effort and errors, increase straight-through processing, and help reinforce compliance measures.

The Ledger Agent supports accounting teams by providing continuous monitoring and insight. It utilises natural-language prompts for monitoring, delivers context-aware explanations with supporting details, and can automatically create and adjust journal entries. Oracle claims this functionality aids in faster issue resolution, lowers the need for handoffs, and improves the accuracy and timeliness of financial visibility.

The Planning Agent assists financial planning and analysis (FP&A) teams. Through real-time trend and variance analysis using natural-language interactions, as well as event-driven predictions on Fusion financial and operational data, it guides decision-makers through what-if scenarios and simulations. The goal is to reduce cycle time for planning processes, improve forecast accuracy, and support more collaborative cross-functional decision-making.

The Payments Agent is developed to help with cash outflow optimisation and payment choices. This agent evaluates options such as early payment programs, virtual cards, and financing, and also enables interactions with bank systems to support faster supplier onboarding and transaction execution.