Microsoft Certified: Azure AI Engineer AssociateImplement image and video processing solutionsMedium
A company is developing an Azure-based solution to automatically detect and redact personally identifiable information (PII) from invoices. The invoices are scanned images containing various layouts and fields. The solution needs to accurately identify names, addresses, and account numbers for redaction. Which Azure AI service is the primary choice for this task?
- AAzure Computer Vision's General OCR
- BAzure Text Analytics for PII detection
- CAzure Computer Vision's Read API
- DAzure Form Recognizer's Prebuilt Invoice model
Show answer & explanationAnswer & explanation
Correct answer: D. Azure Form Recognizer's Prebuilt Invoice model
Azure Form Recognizer's Prebuilt Invoice model is specifically designed to extract structured data, including common PII fields like names and addresses, from invoices with varying layouts. This makes it highly effective for identifying specific information for redaction.
Why the other options are wrong
- A. General OCR extracts raw text but doesn't understand the structure or semantic meaning of PII fields on an invoice.
- B. Text Analytics is for natural language processing of *already extracted text*, not for OCR and structured data extraction from image documents.
- C. The Read API is for robust text extraction (OCR) but doesn't inherently understand invoice structure or identify PII fields semantically.
Azure Form Recognizer Prebuilt Models
Pre-trained models within Azure Form Recognizer that automatically extract structured data from common document types like invoices, receipts, and business cards.
- Understands document structure and semantics
- Extracts key-value pairs and table data
- Reduces need for custom model training for common documents
- Supports various document image formats
Memory trick: Form Recognizer Finds PII on Forms.