Validation checks that your data is correct, complete, and meets your business rules. When a pipeline runs, it generates a validation report detailing any exceptions.
You can review the results, drill down to failed rules, and apply corrections to failed values directly in the report. Every correction is tracked – including who made it and when – providing a complete audit trail and building a library of data corrections to improve data quality over time.
After running a pipeline, go to the Outputs tab and select the Validations tab. This gives you a summary of all validation steps in the pipeline, grouped by stage.
To open the full validation report, click into any validation step.
The report opens with a summary of failed and warned validation rules. Use the filters to show only failed rules, only passed rules, or all rules.
Click into any failed rule to see the rows that failed, along with a summary of the rule at the top of the view.
From here you can:
Next to each column of failed rows is a Corrections column. This is where you apply amendments to the data.
To enter a correction:
If you only want a correction to apply when certain conditions are met, you can add conditions to restrict it — for example, CT>, or only when the policy number contains 1013. Conditions can be stacked for precise control.
Instead of entering a fixed value, you can enter a formula using the calculation editor. Click the FX button (or press =) to open the formula editor — the same formulas available in the calculation and validation editors are supported here.
For example, a formula could add leading zeros to any ZIP code shorter than five digits, or calculate an Effective date from an Expiry date. This is particularly useful when correcting many values without having to type each one individually.
Open the corrections log from the tab in the top right of the rule view. This shows:
From the corrections log you can:
Corrections are applied when the pipeline runs — they are not applied immediately in the report view.
Run the pipeline after making corrections to apply them to your incoming data. Corrections are applied at the point the relevant validation operation appears in the pipeline — data is corrected just before it passes through the validation check.
After the pipeline has run, a blue bar on each rule indicates the proportion of rows that have had corrections applied. Any rows still outstanding are shown separately.
Drilling into a corrected rule shows both the original failed value and the value the pipeline corrected it to — so the process is fully transparent.
Rules where all failed rows have been corrected are shown with a blue checkmark in the Passed filter view.