Extracting Key-Value Pairs from Invoices with Azure AI Document Intelligence

Quick Answer

The answer is the premade invoice model. This is the correct choice because Azure AI Document Intelligence’s premade invoice model is specifically trained to extract key-value pairs—such as invoice number, date, and total amount—from common invoice layouts, handling both PDF and TIFF formats without requiring any custom training. On the Microsoft Azure AI Engineer Associate AI-102 exam, this question tests your ability to distinguish between Document Intelligence models: the layout model extracts text and tables but not key-value pairs, the read model only extracts raw text, and custom extraction requires labeled training data, making the premade invoice model the most efficient for consistent but varied invoice formats. A common trap is confusing the layout model’s table extraction with key-value pair extraction, so remember that key-value pairs are specific to prebuilt models like invoice or receipt. Memory tip: “Invoice for pairs, layout for squares”—the premade invoice model handles the paired data, while layout handles the rectangular tables.