Filter valid, empty, and invalid records from dataset preview

The data preview now supports filtering rows by quality status. This update helps users quickly isolate and understand records that fail validation rules or semantic type checks from the dataset preview, over a 100-rows sample.

With this enhanced preview, you can:

• Filter preview data by status (valid, invalid, empty) to isolate data quality issues.

• Refine filters by specific validation rules or semantic type results to pinpoint the exact cause of invalid data.

• View corresponding indicators that show which rule or semantic type triggered an invalid status.

• Maintain access to an unfiltered sample for overall dataset context.

This improvement enhances the data quality assessment workflow for data engineers, product managers, and quality leads. For example, a data product manager can now preview invalid addresses or missing customer IDs to decide if remediation is needed, while a data quality manager can validate the effectiveness of applied rules in real time.

Filtering is currently available over the 100-row dataset preview sample, with support for a 1000-row sample coming soon to provide even deeper validation visibility.

Configuring data quality