Azure AI-900 Vision Services Guide

Once you identify the visual task, AI-900 may ask you which Azure service family fits it. At fundamentals depth, the main distinction is whether the requirement is broad visual analysis or something specifically face-related.

Service-fit rules

A practical decision rule

Use this rule on AI-900:

• if the scenario is about the image as a whole, start with Azure AI Vision
• if the scenario is specifically about faces, start with the Azure AI Face detection service

If the clue is about receipts, invoices, or forms, remember the broader document-processing boundary from Chapter 1 even though this chapter’s service bullets stay focused on vision and face capabilities.

Common traps

What strong answers usually do

• classify the visual task first
• choose the narrow face service only when the requirement is clearly face-specific
• keep OCR and document reading in the visual lane instead of drifting into NLP or generative AI

Decision order that usually wins

1. Separate broad image analysis, face-specific analysis, and document reading or extraction.
2. If the app needs tags, objects, scenes, or general image understanding, think Azure AI Vision.
3. If the app specifically needs to locate or analyze faces, think Azure AI Face detection service.
4. If the clue is receipts, forms, or line items, reason toward document reading and extraction, not generic image labeling or generative AI.
5. AI-900 often tests service fit by using visual scenarios that sound broader than they really are.

Quiz