Finance AI is moving from experimentation to controlled execution. CFOs, FP&A leaders and EPM administrators are no longer looking only for predictive models or AI-generated summaries. They need AI capabilities that can be assigned securely, governed properly, explained clearly and embedded into day-to-day planning and analysis. 

Oracle Cloud EPM Intelligent Performance Management, or IPM, continues to evolve around this need. The October 2025 Oracle EPM update on Granular Role for IPM and IPM Planning Roadmap focuses on secure finance self-service, AI adoption, Advanced Predictions, Insights, GenAI and the Planning Agent. For finance leaders, the direction is clear: AI in EPM is becoming more accessible to business users, but with stronger role-based control, explainability and governance. 

Product Direction Note 

Oracle includes a safe harbor statement for future-facing product information. Roadmap items, release timing, development plans and pricing may change at Oracle’s discretion and should not be treated as a commitment to deliver specific functionality. 

Organizations should use roadmap content for readiness planning while validating actual availability in their own Oracle Cloud EPM environment. 

The Shift: AI-Powered Finance Inside Oracle Cloud EPM 

Oracle positions AI in Cloud EPM around four finance questions: 

  • Automate: What processes can be fully automated? 
  • Predict: How can finance better anticipate the future? 
  • Monitor: What are the most material exceptions? 
  • Engage: How can GenAI empower finance objectives? 

This is a practical framing because it connects AI directly with finance work. AI is not presented as a separate data science layer. It is embedded into planning, forecasting, analysis, reporting, task automation and business decision support. 

The innovation timeline also shows how Oracle Cloud EPM AI has evolved from Predictive AI in 2016 to AutoPredict, Insights, Predictive Cash Forecasting, Narrative Generation, Modeling Agent, Narrative Summarization and skills-based AI agents and agent teams. This progression matters because finance teams can build AI adoption gradually instead of treating it as a sudden transformation. 

AI Adoption Is Moving Into the Early Majority 

Oracle’s AI adoption curve shows that more than 1,600 Oracle Cloud EPM customers, representing approximately 20%, have at least one AI feature enabled in production. The visual places current adoption near the early majority stage. 

This is an important signal for EPM leaders. AI in finance is no longer limited to a small group of innovators. More organizations are beginning to activate AI features in production environments. The next challenge is not awareness. The next challenge is governance. 

IPM Manage: A Granular Role for Finance Self-Service 

One of the most important updates is IPM Manage, released in 25.10 and available for Enterprise customers. Oracle positions IPM Manage as a granular role that empowers finance teams to configure and run IPM securely without IT bottlenecks. This is a major step for practical AI adoption in finance. 

Without granular roles, organizations often face two extremes. Either AI configuration remains too centralized with Service Administrators, slowing adoption, or access becomes too broad, increasing governance risk. IPM Manage creates a better model by allowing selected finance users to configure IPM jobs while still honoring existing EPM security. 

What IPM Manage Controls 

Oracle identifies IPM Manage as the role for configuring IPM jobs across: 

  • Insights 
  • Auto Predict 
  • Advanced Predictions 

The role supports five governance principles:

Granular Role Control

Finance users can be assigned specific IPM configuration responsibility without receiving broad administrative access.

User and Group Assignment

The role can be assigned to individual users or in bulk through user groups.

Security Honored

Member-level security is respected. Users can only create jobs for members they are allowed to access.

Controlled Job Access

Jobs are accessible only to the creator, while administrators retain full visibility.

Empowered Finance Teams

Finance users can manage IPM configurations with less dependency on IT. 

This creates a stronger foundation for secure finance-led AI. 

Key IPM Manage Use Cases 

Oracle identifies two practical use cases for IPM Manage. 

Individual User 

A business user with IPM Manage can create and manage jobs independently. This reduces reliance on Service Administrators and allows analysts or planners to configure Insights, Auto Predict and Advanced Prediction jobs directly. 

Group Assignment 

The IPM Manage role can be assigned to an FP&A or functional group. Every member of the group can create and manage their own jobs, simplifying administration while enabling the team to operate more efficiently. 

For larger finance teams, group assignment is especially useful. It allows AI job ownership to align with business ownership. 

Why Granular AI Governance Matters 

Finance AI adoption needs speed, but it cannot sacrifice control. A planner may need to configure a forecast prediction job for a specific business unit. An FP&A analyst may need to create Insights for an expense category. A finance manager may need to test Advanced Predictions across revenue drivers. These users should not always need full Service Administrator access. 

Granular roles help balance access and accountability. With IPM Manage, finance teams can adopt AI faster while still protecting member-level security and administrator oversight. This is the right direction for enterprise AI governance: business-led, security-aware and auditable. 

IPM Roadmap: Three Areas of Finance AI Expansion 

Oracle’s IPM roadmap focuses on three major areas: 

  • Insights 
  • Predictive Planning / Advanced Predictions 
  • Gen AI 

The roadmap includes capabilities such as periodic movement variance insights, in-context insights, root cause analysis, Narrative Reporting integration, enhanced Smart View insights, attributes in insights, feature engineering, feature importance, feature selection, what-if predictions, multi-member target selection, non-negative predictions, account-member summarization, non-English GenAI support and Planning Agent. This roadmap shows that Oracle is not only adding AI features. It is improving finance usability, explainability, integration and decision support. 

Advanced Predictions: Making Forecasts Explainable and Actionable 

Forecasting adoption depends on trust. Finance teams will not rely on AI predictions unless they can understand the drivers, challenge assumptions and compare results against business logic. Oracle’s roadmap for Advanced Predictions focuses heavily on explainability and model quality. 

Key areas include: 

  • Feature Importance 
  • Intelligent Feature Engineering 
  • Intelligent Feature Selection 
  • Intelligent Data Selection 
  • Dynamic calc in Advanced Predictions 
  • What-if predictions 
  • Multi-member target selection 
  • Drivers at different POV intersections 

Together, these capabilities help move predictions from statistical output to finance-ready planning intelligence. 

Feature Importance: Explaining Forecast Drivers 

Feature Importance helps finance users understand which business drivers influence forecasts the most. Oracle positions it as a way to compare the relative contribution of each driver, visualize historical data and model estimates using fitted lines, build trust in ML predictions, validate assumptions and highlight key levers of performance. This is critical for FP&A. 

A forecast is only useful when finance can explain why it moved. Feature Importance gives users a way to see whether the forecast is being driven by meaningful business factors such as marketing spend, pricing, inflation, demand indicators, labor cost or other operational variables. 

Intelligent Feature Engineering: Capturing Lag Effects 

Oracle explains feature engineering as the process of preparing data for machine learning by transforming existing features or creating new ones to improve model performance. Supported transformations include time-based features, lag effects, significant lag using autocorrelation, rolling mean and rolling median. This is important because business drivers often affect outcomes with delay. 

For example, marketing spend may not influence sales immediately. A campaign may affect demand over the next three or six periods. Oracle’s visual example highlights marketing spend by period and engineered measures such as marketing spend over the last three and six periods, showing how lag impact can be modeled more effectively. For finance teams, this can improve forecast relevance because the model can reflect how business activity actually behaves over time. 

Intelligent Feature Selection: Removing Noise from Forecast Models 

Too many variables can reduce forecast quality. Oracle states that low-value variables can degrade prediction quality, and feature selection helps choose the most relevant variables by considering how they interact with one another and with the target variable. 

Oracle identifies several algorithms used for feature selection: 

  • Random Forest Regressor 
  • Correlation coefficients 
  • Lasso Regression 
  • Elastic Net 
  • Ridge Regression 
  • Auto Mode 

This improves prediction governance. Instead of including every available driver, the model can focus on variables that add predictive value and filter out noisy or low-impact variables. 

Intelligent Data Selection: Preventing Prediction Job Failures 

Oracle’s short-term roadmap includes Intelligent Data Selection. This capability uses only series with valid actuals in the driver slice for prediction. Series with missing actuals are automatically ignored, helping prevent job failures. Oracle provides an example where future series exist for a Discount % driver, but because actual data is missing, those series are excluded from prediction. 

This is practical for real-world planning environments. Data is often incomplete. Intelligent Data Selection can make prediction jobs more resilient by excluding unsupported series automatically. 

Dynamic Calc in Advanced Predictions 

Oracle identifies Dynamic Calc in Advanced Predictions as a short-term roadmap item. Dynamic calc predictions were already available for Predictive Planning and are being extended to Advanced Predictions. 

This capability allows business users and FP&A teams to make predictions at parent-level or dynamic calc members without depending on granular data. Oracle notes this can help when sub-category data is missing and when users do not want insights at highly detailed account levels. 

A practical example is total Travel & Entertainment, where detailed members such as airfare, hotels, car rental and meals may not provide the best prediction base individually. Parent-level prediction can reduce workarounds and improve usability. 

Insights Roadmap: Bringing AI Directly into Finance Workflows 

Oracle’s Insights roadmap focuses on making AI insights more accessible, contextual and useful. 

Key areas include: 

  • Periodic movement variance insight type 
  • In-context insights 
  • Root Cause Analysis 
  • Insights integration with Narrative Reporting 
  • Enhanced Insights in Smart View 
  • Attributes in Insights 
  • Exclude options in slice definitions 
  • Two years of history in insight graphs 

This direction matters because insights should appear where finance users work — in forms, dashboards, Smart View and reporting workflows. 

In-Context Insights: Surfacing AI Where Users Work 

In-context Insights are designed to make insights available to a wider user base by surfacing them on forms and dashboards. Oracle also positions this as a way to accelerate AI adoption among EPM customers and adapt insight calculations based on custom settings for data intersections. 

This can significantly improve usability. Instead of asking users to visit a separate insights area, EPM can bring relevant exceptions and explanations into the workflow where planning and analysis are already happening. 

Periodic Movement Variance Insight 

Periodic movement variance insight helps users monitor a specific Account member, such as a measure, KPI or metric of interest. It triggers an insight when the change in the recent period is materially significant compared with a benchmark based on the previous period. 

Oracle notes that this will be relevant across Planning, Strategic Workforce Planning, Financial Close, Tax, Profitability and Cost. For finance users, this helps identify material period-over-period movements without manually scanning every account, KPI or measure. 

Root Cause Analysis: Moving from Detection to Explanation 

Root Cause Analysis is positioned for detected insights such as anomalies and forecast variances. It generates a structured breakdown of contributors, potential correlations and causal factors. 

Oracle’s roadmap shows a staged approach: 

  • Driver and contributor analysis at cube level 
  • Causal and correlation analysis at cube level 
  • Causal and correlation analysis using Fusion transactions 
  • Causal and correlation analysis using operational models 
  • Causal and correlation analysis using third-party sources through AI Agent Studio 

This is a major step for finance analytics. Detecting an anomaly is useful. Explaining why it happened is far more valuable. 

GenAI Roadmap: Account-Level Summaries and Planning Agent 

Oracle’s GenAI roadmap includes insight summarization for Account members, GenAI support for non-English languages and Planning Agent. One important enhancement is that multi-insight summaries now support the Account dimension. Previously, users could select multiple insights across dimensions excluding Account in a parent-child context and receive a consolidated summary. Now, users can select insights across multiple account members in a parent-child context to generate a trend summary. 

This is useful because finance teams often analyze performance across account hierarchies such as operating expenses, revenue, cost categories or margin lines. 

Planning Agent: AI Embedded into FP&A Workflows 

Oracle defines Planning Agent as an AI-powered agent embedded into EPM workflows to accelerate planning, analysis and decision-making for FP&A users. Its purpose is to move beyond static reports and manual analysis by delivering proactive, on-demand insights for finance users. 

Oracle also explains why this matters now: enterprises need faster decision cycles, stronger scenario resilience and automated insight generation in an increasingly dynamic business environment. For FP&A teams, Planning Agent represents the next stage of AI in planning. It can help users move from manual review toward guided, contextual and proactive decision support. 

What Finance Leaders Should Prepare For 

The IPM roadmap creates a clear readiness agenda for finance and EPM leaders. 

Organizations should prepare across six areas:

Role Governance

Decide which users or groups should receive IPM Manage and what job ownership model should apply.

Security Review

Validate member-level security so AI jobs respect existing access boundaries.

Forecast Explainability

Prepare users to interpret feature importance, drivers, fitted lines and model assumptions.

Data Engineering Readiness

Review whether business drivers, lag effects, rolling averages and external factors are available for Advanced Predictions.

Insight Adoption

Plan where insights should surface: forms, dashboards, Smart View, Narrative Reporting or executive workflows.

Planning Agent Readiness

Identify FP&A use cases where proactive, on-demand AI assistance can reduce manual analysis and improve decision speed. 

How NexInfo Can Help 

NexInfo helps organizations adopt Oracle Cloud EPM IPM with a practical focus on governance, finance ownership, AI readiness, prediction quality and user adoption. 

NexInfo’s delivery model is supported by ISO 9001 for Quality Management and ISO 27001 for Information Security, helping organizations strengthen process quality, delivery discipline and secure enterprise transformation practices. NexInfo has also received the AI-Enabled Workforce Excellence Award at the 1st Annual Long Beach Business AI Summit, reflecting its focus on practical AI adoption across workforce enablement, enterprise systems and operational transformation.

NexInfo can support organizations with: 

  • Oracle Cloud EPM IPM readiness assessment 
  • IPM Manage role planning 
  • User and group assignment strategy 
  • Member-level security review 
  • Insights configuration and governance 
  • Auto Predict enablement 
  • Advanced Predictions implementation 
  • Feature driver readiness review 
  • Feature Importance adoption planning 
  • Feature Engineering and Selection readiness 
  • Intelligent Data Selection planning 
  • Dynamic calc prediction use case design 
  • In-context Insight adoption strategy 
  • Root Cause Analysis readiness 
  • GenAI insight summarization planning 
  • Planning Agent readiness roadmap 
  • User training and managed support 

NexInfo helps finance teams move from AI experimentation to controlled, secure and measurable AI adoption in Oracle Cloud EPM.  

Conclusion 

Oracle Cloud EPM IPM is becoming more finance-led, more explainable and more embedded in daily planning workflows. The IPM Manage role gives finance teams a secure way to configure AI jobs without broad administrative dependency. Advanced Predictions are moving toward stronger explainability through Feature Importance, Intelligent Feature Engineering, Feature Selection and Dynamic Calc support. Insights are becoming more contextual through forms, dashboards, Smart View, Narrative Reporting and Root Cause Analysis. GenAI and Planning Agent capabilities point toward a more proactive FP&A experience. 

For finance leaders, the opportunity is to bring AI closer to business users while maintaining the governance, security and explainability that enterprise finance requires. 

NexInfo helps organizations adopt Oracle Cloud EPM IPM with Oracle EPM expertise, ISO-certified delivery governance, AI-enabled transformation experience and a finance-first implementation approach. 

FAQ

What is Oracle Cloud EPM IPM?

Oracle Cloud EPM IPM, or Intelligent Performance Management, is Oracle’s AI and machine learning capability for EPM. It supports areas such as predictions, AutoPredict, Insights, narrative capabilities, Advanced Predictions and AI agents.

What is IPM Manage?

IPM Manage is a granular role released in 25.10 for Enterprise customers. It allows selected users to configure IPM jobs such as Insights, Auto Predict and Advanced Predictions.

Why is IPM Manage important?

IPM Manage helps finance teams configure and run IPM securely without depending entirely on IT or Service Administrators. It supports controlled self-service while respecting existing EPM security.

Can IPM Manage be assigned to groups?

Yes. Oracle states that IPM Manage can be assigned to individual users or in bulk through user groups, such as FP&A or functional groups.

Does IPM Manage honor member-level security?

Yes. Oracle states that IPM Manage honors member-level security. Users can create and run jobs only for members they have access to.

What IPM roadmap areas are coming?

Oracle identifies roadmap areas across Insights, Predictive Planning and Advanced Predictions, and GenAI. These include in-context insights, root cause analysis, feature importance, feature engineering, feature selection, what-if predictions, account-member summarization and Planning Agent.

What is Feature Importance in Advanced Predictions?

Feature Importance helps users understand which business drivers influence forecasts the most, compare driver contribution, visualize model estimates and build trust in ML predictions.

What is Intelligent Feature Engineering?

Intelligent Feature Engineering prepares data for machine learning by transforming existing features or creating new ones. Oracle identifies time-based features, lag effects, autocorrelation and rolling mean or median transformations.

What is Intelligent Feature Selection?

Intelligent Feature Selection chooses the most relevant variables from a dataset, removes noisy or low-impact variables and supports explainability by ranking features based on predictive power.

What is Intelligent Data Selection?

Intelligent Data Selection uses only series with valid actuals in the driver slice and automatically ignores series with missing actuals to help prevent prediction job failures.

What are In-Context Insights?

In-Context Insights surface AI insights directly on forms and dashboards, making insights available to a wider user base and helping accelerate AI adoption.

What is Periodic Movement Variance Insight?

Periodic Movement Variance Insight monitors a specific Account member, KPI or metric and triggers an insight when the recent period movement is materially significant compared with a previous-period benchmark.

What is Root Cause Analysis in IPM?

Root Cause Analysis generates a structured breakdown of contributors, potential correlations and causal factors for detected insights such as anomalies or forecast variances.

What is Planning Agent?

Planning Agent is an AI-powered agent embedded into EPM workflows to accelerate planning, analysis and decision-making for FP&A users. It is designed to provide proactive, on-demand insights beyond static reports and manual analysis.

How can NexInfo help with Oracle Cloud EPM IPM?

NexInfo can help with IPM readiness, IPM Manage role planning, security review, Insights configuration, Auto Predict, Advanced Predictions, feature readiness, Root Cause Analysis preparation, Planning Agent roadmap planning, user training and managed support.