AI-103 Certification Practice Question #55

Question

You have a Microsoft Foundry project that contains an agent.
The agent uses Azure AI Search as the retriever.
You plan to ingest PDF into an Azure AI Search index to ensure that the agent can ground responses in texts in both documents and embedded images.
Users require citations that link to the source files.
You need to ensure that during indexing, the images are extracted into a structure that can be used as input for the built-in optical character recognition (OCR) skill.
Which indexing approach should you use?

Accepted Answer

OCR in Azure AI Search runs over normalized images, and that normalized_images collection is produced during document cracking by the **indexer** when `imageAction` is set to `generateNormalizedImages`. MS Learn states outright that the OCR skill assumes a normalized_images field exists and that the indexer's imageAction must generate it — making A the prerequisite indexing step that turns embedded PDF images into a structure the built-in OCR skill can consume. The Shaper skill (B) only reshapes data, OCR against the content field (C) targets text not images, and outputFieldMappings (D) operate after enrichment to surface results, not to extract images. Suggested Answer and MS Learn agree strongly; 88% reflects only the absence of community vote data.

More AI-103 practice questions