At BLUEPRINT 4D 2026, Oracle’s Manish Somani shared a clear vision for the future of enterprise finance: ERP systems are evolving from systems of record into systems of outcomes.
For decades, ERP systems have focused on capturing transactions and storing operational data. But finance teams still spend countless hours extracting reports, reconciling spreadsheets, researching variances, and manually piecing together insights before decisions can be made. According to Somani, AI changes that model entirely.
Oracle Fusion ERP is now being designed to not only record business activity, but also reason over data, generate insights, recommend actions, and increasingly automate execution.
For organizations still running legacy ERP platforms, the message was direct: modernization is no longer just about moving to the cloud. It is about transforming how finance and operations work.
Moving Beyond the Traditional ERP Model
Somani described a major shift happening in enterprise applications. Historically, ERP systems digitized business processes but left analysis and decision-making to humans. Finance users still had to pull data, investigate exceptions, and manually connect information across systems.
Oracle’s approach with Fusion ERP is different. AI is being embedded directly into business workflows so the system can proactively monitor activity, surface anomalies, provide recommendations, and automate repetitive work.
But Somani also cautioned against simply layering AI on top of aging legacy systems. Effective AI-driven ERP requires a strong underlying foundation that understands business rules, security, approvals, hierarchies, and governance.
That distinction between “built-in AI” and “bolt-on AI” will become increasingly important as organizations evaluate modernization strategies. AI can only deliver meaningful outcomes when it operates within the context of trusted enterprise data, established controls, and connected business processes.
Oracle’s Three Pillars of AI-Driven ERP
Oracle’s ERP strategy is built around three key ideas.
The first is machine-led automation. Oracle wants repetitive tasks like data ingestion, invoice processing, and compliance validation handled autonomously wherever possible.
The second pillar focuses on predictive insights. Once data is centralized, AI can identify trends, detect anomalies, and generate recommendations faster than traditional reporting processes.
The third pillar is connected execution. Instead of finance teams analyzing data in one system and collaborating through disconnected tools like email or spreadsheets, Oracle is bringing analysis, collaboration, and action together directly inside the ERP experience.
The goal is simple: reduce friction, speed up decisions, and allow finance teams to focus on higher-value strategic work.
AI-Powered Shared Service Automation
One of the strongest examples shown during the session focused on Oracle’s Payables Agent.
Accounts payable teams often deal with invoices arriving in multiple formats from multiple suppliers. Even with OCR and e-invoicing tools, many organizations still rely heavily on manual review and intervention.
Oracle’s Payables Agent uses AI and large language models to automatically ingest invoices across formats including PDFs, spreadsheets, XML, JSON, and image files without requiring rigid templates or supplier-specific integrations.
The system can validate invoice data, review compliance rules, apply accounting logic, detect anomalies, and even split allocations based on operational context and historical behavior. During the demo, Oracle showed the agent automatically allocating expenses across offices without human involvement.
Somani also highlighted that policy changes can now be updated using natural language rather than requiring technical reconfiguration.
The result is a finance process that becomes faster, more scalable, and significantly less manual.
Reinventing Financial Close with Ledger Agents
Oracle also demonstrated how AI is transforming accounting operations through Ledger Agents.
Traditionally, period close involves extensive manual analysis, report generation, spreadsheet manipulation, and cross-functional collaboration. Oracle’s Ledger Agent changes that experience by continuously monitoring financial conditions in the background.
Users can create natural language prompts to monitor variances, unusual transactions, or operational thresholds without relying on IT-built dashboards.
In one demo, the Ledger Agent investigated a liability variance tied to data center expansion activity. It automatically surfaced invoices, purchase orders, approval comments, and operational context while identifying unfulfilled supply commitments and recommending accrual entries.
Another example showed the system monitoring remaining performance obligations against forecasted revenue and compiling findings into a briefing package ready for finance and planning teams to review collaboratively.
What previously required hours of investigation was completed in minutes through conversational interaction with the ERP system.
Somani also emphasized that Oracle is using these same capabilities internally. Oracle runs its own business on Fusion ERP and has consistently reported quarterly earnings faster than any other company globally for several years — despite operating across more than 150 countries.
Humans Still Matter — But Their Work Changes
One of the more important themes throughout the session was that AI is not intended to replace finance professionals. Instead, Oracle sees AI handling repetitive operational work while humans focus on judgment, strategy, and decision-making.
Tasks like reconciliations, variance analysis, compliance monitoring, and exception handling can increasingly be automated. But finance leaders still provide the context, oversight, and business judgment that AI cannot replicate.
Somani described this as a shift away from checklist-driven work toward higher-value activities that directly impact business outcomes.
Why Clean Data Still Matters
During the Q&A session, Somani emphasized that AI is only as effective as the data and processes behind it.
Organizations moving from legacy ERP platforms cannot simply perform a “lift and shift” migration and expect AI to deliver meaningful value. Modernization requires transformation.
That includes cleansing master data, simplifying chart of accounts structures, standardizing business processes, and eliminating unnecessary customization that does not provide competitive advantage.
Oracle itself went through this process internally before deploying Fusion ERP globally.
The takeaway was clear: standardized processes and clean enterprise data are foundational to successful AI adoption.
The Rise of Agentic Applications
One of the most forward-looking concepts discussed was Oracle’s move toward “agentic applications.”
Somani explained Oracle’s AI evolution in three stages. First came individual AI agents designed for highly specific tasks like document summarization or anomaly detection. Oracle then began connecting those agents into workflows that could manage larger business processes. Now, the company is building “agentic applications” — coordinated teams of AI agents that work together toward broader business goals.
These agents can understand workflows, policies, operational objectives, and historical context while collaborating across finance, HR, supply chain, and customer operations. Instead of simply executing a predefined task, they can increasingly identify exceptions, recommend actions, and guide decision-making.
Oracle is also making its AI Agent Studio available to customers and partners, giving organizations access to the same framework Oracle uses internally to build and manage AI-driven workflows.
ERP Transformation Is Already Underway
The biggest message from the session was that AI is not simply being added to ERP. ERP itself is being redefined.
Finance teams are moving away from spending their time gathering data and managing transactions. Instead, they will increasingly focus on strategy, judgment, and decision-making while AI handles repetitive operational work.
For organizations evaluating their ERP roadmap, Somani’s closing point resonated throughout the session: the question is no longer whether transformation is happening. It is how quickly businesses are prepared to adapt.
Want more?
Explore more content and resources to help you get the most from your Oracle investment:
- Visit the Quest Learn Library for blogs, how-to demos, and on-demand sessions.
- Connect with peers in one of our Quest Cloud Community User Groups to swap stories, ask questions, and share tips with other users facing the same challenges.
Not a Quest member yet? Join today and tap into the ultimate Oracle customer network.