ReportingIQ Add-on: To utilize AI-powered similarity scoring for duplicate detection, the ReportingIQ Paid-Tier add-on must be installed. Without it, the Duplicate Review features will not be available, but other features are still enabled
​
​

Permissions: Users must have the necessary permissions to view the Data Hygiene Hub and edit the respective core tables they are cleaning up.

Use the Core tables menu on the left sidebar to select the data set you want to review (e.g., Customers, Products, Brands, Vendors).
​
​

The dashboard cards will display the current status of the selected table, allowing you to jump directly into the cleanup workflow that matches the issue type: Potential duplicates, Rows with gaps, or Inactive candidates/Unused records.

Navigate to the Duplicate review tab.
​
​

Select a match group from the Duplicate clusters list on the left.
​
​

In the Resolve duplicate records panel, review the flagged records. The system will automatically suggest a recommended master record.
​
​

Choose which record should be kept as the master by selecting the radio button next to its name.
​
​

For any fields with conflicting data (marked with a warning icon), verify the data you want to keep.
​
​

Click Merge into [Master Record Name] on the duplicate records to merge their history, notes, and transactions into the master record.

Navigate to the Missing data tab.
​
​

Review the grid containing records missing critical information. Cells highlighted in yellow indicate missing required data.
​
​

Click directly into the table cells to type, select from dropdowns, or use autocomplete to fill in the missing information.
​
​

Check the boxes on the left side of the rows you have updated (or use the select-all checkbox at the top).
​
​

Click the Save changes button in the top right corner to apply all your edits at once.

Navigate to the Stale records (or Unused records) tab.
​
​

For Customers and Products (Stale Records): Select records that have been idle past your configured freshness threshold, then click Inactivate selected records to change their status to inactive without deleting data.
​
​

For Lookup Tables like Brands, Vendors, etc. (Unused Records): Review records with no linked sub-table activity. If the system confirms a record has 0 linked records, you can click the red Delete button to permanently remove it from the system.

Similarity status: A panel indicating whether the AI similarity scoring is currently active and running on the selected table's search field.
​
​

Match confidence: A percentage score indicating how confident the AI is that the clustered records are duplicates.
​
​

Fields with differences: A count of how many data fields differ between the records in a duplicate cluster.
​
​

Recommended Master: The specific record within a duplicate cluster that the system flags as the most complete or relevant record to keep during a merge.
​
​

Linked records: Found in the Stale/Unused tab, this number shows how many transactions or sub-table entries are tied to the record, helping you decide if it is safe to delete.

Review Before Merging: Similarity scoring can occasionally produce false positives. Always manually review the field differences in a duplicate cluster before committing to a merge.
​
​

Save Frequently: When working in the Missing data grid, save your changes in batches rather than trying to update hundreds of rows at once.
​
​

Check Linked Records: Never delete an unused record without confirming that the Linked records count is 0.