An Episode and Article in the EPM Intelligence Oracle AI World Series featuring Marc Seewald (Oracle VP of Product Management)
Key Takeaways
-
Oracle’s AI planning journey is evolutionary: progressing from deterministic and predictive models to generative and now agentic AI over nearly a decade.
-
Trust is foundational for AI in finance, with explainability, auditability, and deterministic checks embedded directly into planning workflows.
-
The Planning Agent encourages CFOs and executives to interact directly with scenarios, insights, and forecasts.
-
Operational data improves forecasts: Connecting sales and operational plans to finance, gives AI the leading indicators it needs to perform and improve analysis.
The Evolution of Oracle AI Planning: A Timeline
2016: Deterministic & Predictive Foundations
Oracle releases its first AI features in planning, focused on predictive forecasting using historical data and forward-looking signals. These capabilities support FP&A as well as operational planning use cases.
“We released our first AI feature back in 2016. So this is going on almost ten years now.”
Predictive Era: AI-Driven Insights for Finance
Building on predictive forecasting, Oracle introduces AI-powered insights designed to automatically detect variances, trends, and anomalies—reducing manual analysis and accelerating understanding for finance teams.
“You can think about it as a pattern detection engine built for finance.”
2024: Generative AI in Planning
Oracle introduces generative AI into narrative reporting and insights, enabling automated explanations and summaries while maintaining transparency and human oversight.
Most Recent: Agentic AI and the Planning Agent
Oracle introduces agentic AI through the Planning Agent, combining predictive, generative, and deterministic AI into a conversational experience that helps users analyze, model, and act on planning data.
Ensuring Transparency In Critical Financial Planning Answers
Finance demands a higher standard of trust than most other domains, and Oracle’s AI strategy reflects that reality.
Rather than relying solely on probabilistic models, Oracle applies the right AI technique to the right task. Generative AI may be used for narratives and summaries, but numerical outputs are validated through deterministic calculations and rules behind the scenes.
For predictive forecasts, Oracle emphasizes explainability. Users can see:
- Historical anomalies that influenced results
- Which drivers had the greatest impact on a forecast
- How multiple variables interacted to produce an outcome
For generative outputs, Oracle clearly labels AI-generated content and is building auditability into how prompts are interpreted and responses are created. The goal is not to turn finance users into data scientists, but to make AI decisions understandable and defensible.
The Planning Agent And Its New User, The CFO
One of the most significant shifts introduced by agentic AI is who can now use planning systems.
Historically, senior executives relied on finance teams to run scenarios, investigate variances, and prepare management materials. The Planning Agent changes this dynamic by enabling conversational interaction with planning data.
With the agent, users can:
- Ask follow-up questions on variances and root causes
- Perform scenario and “what-if” analysis using natural language
- Automatically generate visualizations and narratives
- Push insights directly into management reporting packages
This lowers the barrier to entry for executives, particularly CFOs, allowing them to explore scenarios and strategic questions directly, without navigating complex planning interfaces
Traditional Planning Features To Get Started With
While AI is a powerful accelerator, Marc emphasizes that strong planning fundamentals come first.
“Even before you get into AI, there’s all kinds of things you can do to automate your process.”
Many organizations still rely on spreadsheets for budgeting, forecasting, and allocations. Even without AI, modern planning platforms offer immediate value through:
- Rolling forecasts
- Automated allocations
- Narrative reporting
- Reduced manual data entry
These capabilities not only improve efficiency on their own, they also create the structured, connected data foundation required for AI to deliver meaningful results.
Planning Outside of Finance
Planning is no longer a finance-only discipline. Sales, supply chain, IT, and other operational teams hold assumptions that directly shape financial outcomes—often outside formal planning processes.
As Marc explains, the distinction is fundamental:
“Financial data tends to be the lagging indicator, and the operational data tends to be the leading indicator.”
When operational and financial planning are connected:
- Forecast accuracy improves through the use of leading indicators
- AI models become more effective with richer, more relevant inputs
- Risks and opportunities surface earlier, increasing business agility
Sales planning illustrates the point. CRM opportunities alone are not forecasts; judgment, probability, and timing matter. When these operational assumptions feed financial planning, AI produces insights that are more realistic, timely, and actionable.
AI Planning Leading to Greater Focus
Marc shares a customer example that illustrates how AI changes where finance teams spend their time, not just how fast they work.
“There’s a customer … a large Fortune 100 company, and they had rolled out our AI features a number of years ago.”
The company deployed AI-driven predictions and insights across five major business units. Four of those units saw significant gains in efficiency, forecast accuracy, and insight generation.
The fifth unit told a different story, and that difference mattered.
As Marc explains, the CFO viewed the uneven impact as a success, not a failure. The most complex, high-touch business unit benefited less from automation, which allowed finance leaders to reallocate resources away from routine analysis and toward the areas where judgment and partnership were most critical.
The result was not fewer finance professionals, but better focus. Spending less time on repeatable work and more time supporting complex decisions. This is the real value of AI in planning: amplifying human expertise by freeing it from lower-value tasks.
EPMI’s Key Insights
AI, EPM, and ERP are reshaping enterprise performance by moving intelligence directly into the systems where decisions are made. Modernizing planning isn’t about chasing generative AI headlines, it’s about improving forecast accuracy, increasing trust in the numbers, and enabling leaders to respond faster to change.
When planning is connected, continuously optimized, and intelligently automated, it becomes a strategic advantage rather than a reporting exercise.
At EPMI, we help organizations start with clear goals, unify their data, and empower leaders to make better business decisions using Oracle’s built-in AI and enterprise solutions.
FAQs
1. How is Oracle’s AI in planning different from generic generative AI tools?
Oracle embeds AI directly into planning workflows, combining predictive, generative, and deterministic capabilities with governance, security, and financial controls that generic tools lack.
2. Can organizations use AI in Oracle EPM without being advanced in their planning maturity?
Yes. Oracle encourages starting small, often with predictive forecasting or insights, while building foundational planning processes that AI can later enhance.
3. How does Oracle ensure AI-generated forecasts can be trusted?
Through explainability for predictions, deterministic validation for numbers, and clear labeling and auditability for generative outputs.
4. Who benefits most from the Planning Agent?
CFOs and senior leaders gain the most immediate value, as the agent enables direct interaction with scenarios, drivers, and insights without technical intermediaries.
5. Why is operational planning so critical to AI-driven forecasting?
Operational data provides leading indicators of business performance; without it, AI forecasts rely too heavily on lagging financial results and lose predictive power.
Watch the AI World Series on YouTube!