Question
How can I match the data in ‘Documents Classified’ with ‘Human in the Loop’ on the VMR dashboard?
Answer
To reconcile the data displayed in the VMR dashboard under the ‘Documents Classified’ and ‘Human in the Loop (HITL)’ menus, users need to understand how documents are processed and categorised—either automatically or through manual intervention. This helps ensure transparency in automation performance and allows users to track which documents are ready for export and which require further action.
Scenario Overview
Between 1 July and 14 July 2025, a total of 46 documents were loaded into VMR, as shown in the image below.
Reconciliation Breakdown
- Automatically classified: 46 documents
- Import error: 0
- Automation is disabled for file, but it remains accessible in the “Loaded” section of the VMR dashboard, along with its error message, when a document loaded into VMR encounter technical issue.
- Documents ready for export: 26
- These are listed under the “Ready” menu and can be exported to Class Fund Documents or SharePoint.
- Published: 18
- Documents requiring human intervention: 2
- 2 documents flagged as Errors
- These documents are routed to the HITL queue for manual review and correction by VMR users.
User Perspective
From a user’s perspective, reconciliation involves:
- Reviewing the ‘Documents Classified’ section to identify files that have been successfully processed and are ready for export.
- Accessing the ‘HITL’ section to locate documents that require manual updates, such as correcting document types or mapping data points.
- Ensuring all flagged documents are reviewed and resolved to maintain data accuracy and workflow continuity.