NetSuite EPM AI Availability and Setup Requirements

Compare 16 NetSuite EPM AI features across seven business processes, reconciliation setup requirements, and the January 2027 Generative AI deadline.

·5 min read·12 views·NetSuite 2026.2·View Oracle Docs

NetSuite EPM lists 16 AI capabilities across seven business processes, from reconciliation predictions to reporting assistance. Availability depends on business process, environment update, region, and account configuration. Generative AI has a specific prerequisite: environments must reach Cloud EPM update 26.08 or later by the end of January 2027 to retain access.

What Does the EPM AI Overview Cover?

The documentation brings together 16 AI capabilities and identifies which EPM business processes support them. Treat this as an availability reference, not a list of features introduced together in one release. The accompanying details explain two Account Reconciliation features and their different configuration requirements.

For broader assistant context, see NetSuite EPM AI assistants and financial insights. Start here by identifying your business process, checking the applicable features, and verifying environment prerequisites before planning adoption.

Which AI Features Are Available for Each Business Process?

The table below preserves the documented availability mapping. Each row lists every business process marked as supported for that feature; processes omitted from a row are not marked as supported in the matrix.

AI featureSupported EPM business processes
Reconciliation Assignment AssistanceAccount Reconciliation
Transaction Matching AssistanceAccount Reconciliation
PCM AgentProfitability and Cost Management
Consolidation AssistantClose Management and Consolidation
Generative AI in Data DiscoveryClose Management and Consolidation
Consolidation Job AnalyticsClose Management and Consolidation
Advanced PredictionsFreeForm; Planning and Budgeting
Auto PredictFreeForm; Planning and Budgeting
Predictive PlanningFreeForm; Planning and Budgeting
Bring Your Own MLFreeForm; Planning and Budgeting
Monte Carlo SimulationPlanning and Budgeting
Notes Summarization with Generative AINarrative Reporting
Generative AI for Management Reporting NarrativesNarrative Reporting
Return to Accrual Automation AI AssistantTax Reporting
IPM InsightsProfitability and Cost Management; Close Management and Consolidation; FreeForm; Planning and Budgeting; Tax Reporting
Reporting AgentAccount Reconciliation; Profitability and Cost Management; Close Management and Consolidation; FreeForm; Planning and Budgeting; Narrative Reporting; Tax Reporting

Reporting Agent is the only listed capability marked available across all seven processes. IPM Insights spans five. Planning and Budgeting shares four prediction and machine-learning capabilities with FreeForm, while Monte Carlo Simulation is listed only for Planning and Budgeting.

Availability is conditional: Business process, environment update, region, and account configuration can all affect access. A supported entry is therefore a starting point for validation, not a guarantee that the feature is immediately usable in every environment.

What Is the Generative AI Upgrade Deadline?

Generative AI requires the August 2026 Cloud EPM update, identified as 26.08, or a later update by the end of January 2027. Environments that have not reached that update level by the deadline will lose access to Generative AI features.

Administrators should check the actual Cloud EPM environment update when assessing this requirement. As a rollout precaution, inventory the generative features your teams use and verify access after updating. The documentation does not extend this loss-of-access warning to every predictive or machine-learning capability in the table.

How Does Reconciliation Assignment Assistance Work?

Reconciliation Assignment Assistance predicts a specified attribute value using machine-learning algorithms trained on historical data. Service Administrators create a predicted attribute and train it against existing data. Training produces a reusable model that can generate predictions any number of times and can be retrained periodically as data changes.

Each predicted value has an associated confidence-percentage attribute. Higher confidence indicates a prediction is likely more accurate. The documented benefits include less manual work creating profile attributes, faster period-end close, and greater consistency and accuracy through fewer human errors and historical-data-driven insights.

No enablement action is required, but preparation is still necessary. Service Administrators must create and train the predicted attribute before using its model. Distinguish feature enablement from model readiness when assigning implementation work: no enablement action does not mean there is no setup work.

A useful validation step is to review predictions alongside their confidence percentages and assess whether the historical training data still represents current practice. Periodic retraining is supported; the documentation does not prescribe a retraining interval or confidence threshold.

How Does Transaction Matching Assistance Work?

Transaction Matching Assistance predicts matches using a model trained on historical manual-match data. Users can review each predicted match and confirm or discard it. A confidence score accompanies each prediction, with higher scores indicating more reliable matches and helping users prioritize their manual review.

Oracle describes Auto Match as handling the majority of loaded transactions, leaving exceptions for manual matching. Transaction Matching Assistance predicts potential matches for unmatched transactions. Its documented benefits are reduced manual matching effort, a faster period-end close, and more accurate, consistent predictions based on historical patterns.

To use it, select Enable Predictive AI in the configuration settings. Unlike Assignment Assistance, this feature has an explicit enablement requirement. The documentation provided references Configuring Artificial Intelligence Settings but does not specify a navigation path, training procedure, or required confidence cutoff.

For related integration work, consult the Account Reconciliation Sync SuiteApp enhancements. Keep sync verification and predictive-feature validation as separate checklist items rather than assuming one confirms the other.

Who This Affects

  • Service Administrators: Create and train predicted attributes for Assignment Assistance, plan retraining, and verify applicable AI configuration and environment readiness.
  • Reconciliation teams: Review matching predictions, use confidence scores to prioritize attention, and confirm or discard proposed matches.
  • Planning, consolidation, profitability, reporting, and tax teams: Use the availability matrix to scope evaluations for their own business processes.
  • EPM environment owners: Track the Generative AI update deadline and validate continued access after updates.

Key Takeaways

  • The matrix covers 16 AI features across seven EPM business processes; Reporting Agent spans all seven.
  • Generative AI access requires Cloud EPM 26.08 or later by the end of January 2027.
  • Assignment Assistance needs predicted-attribute creation and training, despite requiring no enablement action.
  • Transaction Matching Assistance requires Enable Predictive AI and lets users confirm or discard predictions.
  • Validate business process, environment update, region, and account configuration before committing to feature availability.

Frequently Asked Questions (4)

Which AI features are shared by FreeForm and Planning and Budgeting?
Both processes list Advanced Predictions, Auto Predict, Predictive Planning, and Bring Your Own ML. Monte Carlo Simulation is listed only for Planning and Budgeting.
What Cloud EPM update is required to retain Generative AI access, and what happens if an environment misses the deadline?
The environment must reach Cloud EPM update 26.08 or later by the end of January 2027. Environments that miss the deadline will lose access to Generative AI features; the article does not extend this warning to all predictive or machine-learning capabilities.
Does Reconciliation Assignment Assistance need to be enabled, and what setup is required before using it?
No enablement action is required, but a Service Administrator must create a predicted attribute and train it on existing historical data. Training creates a reusable model that can generate predictions repeatedly and be retrained as data changes.
How do users review Transaction Matching Assistance predictions, and what must be enabled first?
Select **Enable Predictive AI** in the configuration settings. Users can then review predicted matches for unmatched transactions, use each prediction’s confidence score to prioritize review, and confirm or discard the proposed match.

Weekly Update History (1)

NetSuite Enterprise Performance Managementadded

Added an entry for NetSuite EPM Features with AI.

View Oracle Docs
Source: NetSuite EPM Features with AI — Oracle NetSuite Help Center. This article was generated from official Oracle documentation and enriched with additional context and best practices.

Was this article helpful?

More in AI

View all AI articles →

Also from NetSuite 2026.2

View all NetSuite 2026.2 changes →