Domain 2 of 2 · Chapter 4 of 4

Implement AI solutions for information extraction by using Foundry

Unlock the complete study guide + 1,040 practice questions across 16 full exams.

Bundled into the existing Microsoft Azure AI Fundamentals premium course — no separate purchase.

14-day money-back guarantee — no questions asked.

Included in this chapter:

  • An analyzer, and the two passes it runs
  • Where your analyzer comes from
  • Writing the field schema
  • Documents and forms
  • Images: reusable fields, not a one-off answer
  • Audio and video: built on the transcript
  • Calling an analyzer from an application
  • Exam patterns for information extraction

What changes across the four analyzer modalities

Decision pointDocumentImageAudioVideo
Base analyzer (baseAnalyzerId)prebuilt-documentprebuilt-imageprebuilt-audioprebuilt-video
What content extraction returnsOCR text, layout, paragraphs, tables, selection marks, barcodes and formulasThe image for the model to reason over; text-heavy images are better served by a document schemaA speaker-separated WebVTT transcript, with optional speaker-role detectionA WebVTT transcript plus ordered key frames and shot boundaries on one timeline
Configuration options that applyreturnDetails, omitContent, enableOcr, enableLayout, tableFormat, enableSegment, segmentPerPage, among othersreturnDetails and disableFaceBlurring onlyreturnDetails and localesreturnDetails, locales, contentCategories, enableSegment, omitContent, disableFaceBlurring
Field methods you can writeextract, classify and generateclassify and generate (extract is documented for documents)classify and generate (extract is documented for documents)classify and generate (extract is documented for documents)
Prebuilt analyzer to try firstprebuilt-invoice, prebuilt-receipt, prebuilt-idDocument, prebuilt-layout, prebuilt-documentSearchprebuilt-imageSearchprebuilt-callCenter, prebuilt-audioSearchThe prebuilt retrieval video analyzer, which Microsoft's video overview names two different ways on one page
Telltale requirement"pull these fields off every incoming form""describe every product photo for the catalog index""summarize each support call and score its sentiment""split the broadcast into stories, each with its transcript"

Decision tree

Are the values alreadyin the file?Deploy a generative modelcreates content, not extractionNoYesSame fields from everyfile, in a fixed shape?Multimodal chat requestone answer, not a schemaNoYesWhich modality isthe input?DocumentsImagesAudioVideoDocument analyzerprebuilt-documentprebuilt-invoiceImage analyzerprebuilt-imageprebuilt-imageSearchAudio analyzerprebuilt-audioprebuilt-audioSearchVideo analyzerprebuilt-videotranscript and segmentsEvery analyzer call is asynchronousbegin_analyze then poller.result(), or 202 Accepted then poll Operation-Location
Cheat sheet

Cheat sheet

  • An analyzer defines content processing and output
  • Analyzer schemas produce machine-readable fields
  • Document analyzers extract layout and structured fields
  • A prebuilt invoice analyzer handles common invoice fields
  • Image analyzers extract structured visual insights
  • Prebuilt image analysis returns a structured summary field
  • Audio analyzers return transcripts and audio-specific insights
  • Video analyzers combine timed visual and audio evidence
  • ContentUnderstandingClient connects with an endpoint and credential
  • begin_analyze returns a poller for asynchronous analysis
  • AnalysisInput can reference content by URL
  • REST analysis polls the Operation-Location URL
  • Analysis results separate normalized content from extracted fields

Unlock with Premium — includes all practice exams and the complete study guide.

Also tested in

References

  1. Azure Content Understanding in Foundry Tools - What is an analyzer? Configuration and reference
  2. What is Azure Content Understanding in Foundry Tools?
  3. Azure Content Understanding in Foundry Tools document overview
  4. Azure Content Understanding in Foundry Tools prebuilt analyzers
  5. Azure Content Understanding in Foundry Tools image overview
  6. Azure Content Understanding in Foundry Tools audio overview
  7. Azure Content Understanding in Foundry Tools video overview
  8. Quickstart: Azure Content Understanding in Foundry Tools REST API and SDKs