AI-901 ยท Common AI workloads
What does Azure Content Understanding in Foundry Tools do?
Pulling structured data out of text, images, audio and video: fields from forms and invoices, transcripts, scene and segment descriptions, and when extraction beats plain chat.
Key points
- Processes documents, images, audio and video through an analyzer (prebuilt or custom).
- Output: Markdown for search and retrieval, or JSON fields matching your schema.
- Adds confidence scores and grounding to the source.
- Uses: invoice fields, call transcripts and insights, video split into scenes with descriptions.
- Use it instead of plain chat for consistent, verifiable structure at scale.
Source: https://learn.microsoft.com/en-us/azure/ai-services/content-understanding/overview
Practice questions
Which Foundry tool turns documents, images, audio and video into output that follows a schema you define?
- Azure Speech in Foundry Tools
- Azure Content Understanding in Foundry Tools
- Azure Language sentiment analysis
- A text-to-image generation model
Show answer
Answer: B. Azure Content Understanding in Foundry Tools
Azure Content Understanding in Foundry Tools applies analyzers to many content types and returns structured output. Azure Speech covers audio only, and Language covers text analysis.
Which outputs can Content Understanding provide?
- Markdown representation of the content
- Structured JSON fields matching your schema
- A confidence score for each extracted field
- An automatically deployed customer web app
- A retrained base language model after every request
Show answer
Answer: A. Markdown representation of the content, B. Structured JSON fields matching your schema, C. A confidence score for each extracted field
It can return Markdown, JSON fields matching your schema, and confidence scores per field. It does not deploy a web app for you or retrain a base model as part of an analysis call.
True or false: Content Understanding can only process documents, not audio or video.
Show answer
Answer: False
False. It handles documents, images, audio and video; for example it can transcribe calls or divide a video into scenes.
Practice 8 questions on this topic in Bitzu
Short lessons, instant explanations and spaced review of what you get wrong. Free on iOS and Android.
