AI-900 Practice Question: Describe features of computer vision workloads on Azure

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Azure Face API

  Why it's wrong here

  Azure Face API is incorrect for this task because it is a specialized service for detecting, analyzing, and recognizing human faces in images. It performs operations like facial attribute extraction, emotion detection, and face verification, but it has no OCR capabilities to read or extract text from documents. Scanned invoices contain no faces, so this service would be completely unresponsive to text extraction requirements.

  When this WOULD be correct

  A question asking which Azure service can detect and identify individuals in a photo or video feed, such as for security or attendance tracking, would make Face API the correct answer.

• ✓

  Azure AI Document Intelligence (formerly Form Recognizer)

  Why this is correct

  Azure AI Document Intelligence is the correct service because it is purpose-built for extracting text, including hand-written characters, and structured data from documents like invoices. It combines optical character recognition (OCR) with prebuilt domain-specific models that parse key-value pairs, tables, and line items from scanned invoices, going beyond raw text to deliver semantic understanding. This service directly meets the requirement to automatically extract text from scanned invoices.

• ✗

  Azure Custom Vision

  Why it's wrong here

  Azure Custom Vision is incorrect because it is a model-training service for image classification and object detection; it enables users to build custom models that classify entire images or identify object locations, but it does not perform text extraction. It has no built-in OCR capability and cannot read or understand the content of a scanned invoice. Any text extraction would require an external OCR engine, making this service unsuitable as the primary solution for invoice text extraction.

  When this WOULD be correct

  A company needs to classify scanned invoices into categories (e.g., 'paid' vs 'unpaid') based on visual features like logos or stamps, without extracting text content.

• ✗

  Azure Video Indexer

  Why it's wrong here

  Azure Video Indexer is incorrect because it is designed to process video content, extracting insights such as spoken language via speech-to-text, on-screen faces, and scene descriptions. Although it can perform OCR on text visible within video frames, it is not meant for static scanned invoices and lacks the invoice-specific layout analysis and structured data extraction that document-focused OCR tools provide. Using it would require uploading invoices as video, which is neither practical nor optimized for this scenario.

  When this WOULD be correct

  A question asking which Azure service extracts metadata, transcripts, and spoken content from recorded meetings or video files would make Azure Video Indexer the correct answer.

Option-by-option analysis

Why each answer is right or wrong

Understanding why wrong answers are wrong — and when they would be correct — is what separates a 750 score from a 900. The AI-900 exam frequently reuses these exact scenarios with slightly different constraints.

Analysis generated from the official AI-900blueprint and verified against question context. The “when correct” sections are what AI assistants cite when candidates ask “what’s the difference between these options?”

Go deeper

Related to this question