The PII/PHI Review workflow requires a dataset containing the documents to be analyzed for sensitive information.
When preparing input data for this workflow:
- Import a dataset that reflects the scope of documents under review.
- Ensure the documents contain extracted text. The workflow analyzes document text, so documents that failed text extraction cannot be reviewed and will appear on the exception report.
- Apply appropriate filtering before import to limit the dataset to relevant documents.
Breach populations often include scanned forms, image-only PDFs, and photographs of documents — exactly the material most likely to carry identifiers. Confirm text extraction coverage before committing a full population.
Steps
The Steps tab lists the components used to detect sensitive information and identify the individuals affected. Each component represents a step in the analysis and contributes to the overall workflow outcome.
In this workflow, the components are:
| Component | Purpose |
|---|
| PII Detection | Determines whether each document contains PII or PHI, tags the specific elements found, and captures the supporting text, a confidence level, and a count for each one. |
| PII Entity Extraction | Identifies the individuals the detected information belongs to and records their contact details and the elements exposed for each of them. |
The available configuration options depend on the component. Components may require you to:
- Review and edit the element definitions that decide what is reportable
- Confirm the tags that correspond to those elements
- Set model options such as the classification threshold
- Select the output columns the component writes
Configuration changes affect how the workflow evaluates documents, so review each component carefully before running the workflow.
PII Entity Extraction depends on PII Detection. Extraction only processes documents that carry a determination from detection, and it treats that determination as fixed. Run detection first, review it, and then run extraction — and after any correction, re-run both in that order so the document population and the individual population stay consistent.
Step Results
The Results tab shows how the PII/PHI Review workflow progresses through its components once it is run.
The Steps tab is also where you review results. Each component opens to its own views:
- PII Detection — a Results view listing every document with its determination, the elements found, the supporting text behind each element, and the model score; and a Metrics view reporting how the model is performing against reviewed documents.
- PII Entity Extraction — a Documents view, an Entities view with one row per individual, and a Document entities view with one row per document-and-individual pair.
Use the Steps tab to monitor execution and to validate results before relying on them.
Outputs
The Outputs tab displays the consolidated results generated by the PII/PHI Review workflow.
Outputs may include:
- Document-level determinations and the elements detected in each document
- The affected-individual population, consolidated across the document set
- Reports summarizing what was found and accounting for what could not be processed
- Exported tables for downstream reporting and notification work
Two reports are available:
| Report | Purpose |
|---|
| Prediction output | The headline numbers for the review — documents reviewed, documents containing PII, documents containing PHI, affected individuals, and element types detected — with a breakdown of each sensitive data element by the number of individuals and documents it appears in. |
| Exceptions | A reconciliation of documents submitted against documents that produced a result, with the shortfall broken down by cause and a recommended action for each. |
Both reports can be exported to Excel or PDF. The entity table, the document-entity table, and the per-document results can be exported to CSV.
Review outputs to validate results and determine next steps in your review process.
Runs
The Runs tab provides a history of workflow activity.
From this tab, you can:
- See when the PII/PHI Review workflow was run or deployed
- Review the status of past runs
- Identify which runs completed successfully and which may require attention
- Confirm that detection ran before extraction following a correction
Run history helps support tracking, auditing, and troubleshooting over the lifecycle of the workflow.