GenAI-based reconciliation summary

The Data Ledger includes an AI-powered feature that provides human-readable summaries of reconciliation data across your tenant. Infor GenAI is used to analyze 30 days of reconciliation data, identify mismatches between source applications and Data Lake, and present findings in a structured, actionable format.

Reconciliation data is summarized by GenAI at these levels:

  • Quick summary: An overview of the reconciliation health, mismatch count, and recommended investigation status is displayed on the Data Ledger page.
  • Detailed summary: A detailed breakdown by scope, time window, affected data objects, and ingestion type (batch or streaming) that is available in a side panel.

Each summary generates a structured report that includes this information:

  • Reconciliation health status: Healthy or unhealthy
  • Scope details: Source application, ingestion type, and time window
  • Mismatch counts
  • Mismatch occurrence timing
  • Affected data objects
  • Contextual notes on severity and continuity patterns

All GenAI summaries include a disclaimer indicating that the content is AI-generated and can contain inaccuracies. We recommend that you verify critical information before you act based on AI-generated recommendations.

GenAI summarization enforces tenant isolation. Only reconciliation data from your tenant is analyzed. No prompts or AI responses are retained beyond the request.

Quick summary

When the GenAI feature is enabled for your tenant, a short AI-generated summary is displayed in the Data Ledger page header. This automatically-generated summary provides an overview of the reconciliation health across all source applications.

The quick summary includes information on the overall reconciliation health, total number of mismatches detected in the last 30 days, and a note whether an investigation is required.

Detailed summary

To access the detailed GenAI reconciliation analysis, click the Arrow icon or the View full report link in the Data Ledger page header. A side panel is displayed with the structured report for the last 30 days and includes these elements:

  • Scope: Source application, ingestion type (batch or streaming), and analysis time window.
  • Reconciliation status: Presence and count of mismatches.
  • Mismatch occurrence timing: Grouped-by-hour time periods with the highest mismatch counts.
  • Affected data objects: Data objects most frequently affected by mismatches.
  • Notes: Contextual information on the health, continuity patterns (continuous failure or intermittent issues), and dismissed mismatches, if applicable.

To refresh the analysis with the latest reconciliation data, click Regenerate. To close the panel, click Close.