Microsoft Certified: Azure AI Engineer AssociateImplement knowledge mining solutionsHard
A healthcare provider is implementing an Azure AI Search solution to index patient records stored in Azure Blob Storage. These records are primarily in PDF format and contain sensitive health information. The solution needs to redact specific personally identifiable information (PII) like patient names and social security numbers before indexing. Which type of custom skill should be developed and integrated into the indexer's skillset to achieve this redaction?
- AA custom Azure Function skill implementing redaction logic.
- BA custom Azure Machine Learning skill for sensitive data classification.
- CA custom web API skill calling an external PII detection service.
- DA built-in Entity Recognition skill configured for PII.
Show answer & explanationAnswer & explanation
Correct answer: A. A custom Azure Function skill implementing redaction logic.
While built-in entity recognition can detect PII, it primarily extracts; it does not redact. To perform actual redaction (modifying the content before indexing), a custom skill implemented as an Azure Function or a custom web API is required. An Azure Function is a common and efficient way to host such custom logic within a skillset.
Why the other options are wrong
- B. While ML could classify sensitive data, the requirement is for redaction before indexing, which is an operational task best handled by a custom skill in the indexing pipeline, not just classification.
- C. A custom web API skill could work, but an Azure Function is often a more integrated and cost-effective solution for custom logic within Azure.
- D. Built-in Entity Recognition skills can detect PII, but they do not perform redaction; they extract the entities. Redaction requires modifying the content.
Custom Skill for Redaction
An Azure AI Search custom skill, typically implemented as an Azure Function or Web API, used to modify document content (e.g., redact PII) during the indexing pipeline.
- Allows for arbitrary code execution.
- Can be chained with other built-in or custom skills.
- Essential for data transformation or redaction not covered by built-in skills.
Memory trick: To hide secrets, you need a 'custom function' to 'scrub' the data clean before anyone searches.