Chapter 3 of 7 · Microsoft AI-901
Identify AI Workloads
Generative and agentic AI, text analysis, speech, computer vision, image generation and information extraction.
- 3topics
- 14 minfast track
Topics in this chapter
Take them in order, or jump to the one you need.
- 1AI Workloads and Text AnalysisAI-901 AI workloads: generative AI, agentic AI, text analysis, speech, vision and extraction, plus key phrases, entities…5 min
- 2Speech and Computer Vision CapabilitiesAI-901 speech and vision: speech recognition and synthesis, real time and batch transcription, SSML, image analysis, OCR…5 min
- 3Extracting Information from ContentAI-901 information extraction: OCR and layout, fields with confidence scores, entity extraction, transcripts from audio, and…4 min
About this chapter
What this chapter covers
An AI workload is a kind of job that AI does. This chapter introduces the main workloads, generative and agentic AI, text analysis, speech, computer vision and image generation, and information extraction, and the techniques each one uses.
Questions describe a business need and ask which workload or technique fits. The skill is recognising the shape of the problem.
How the topics fit together
Each workload works on a different kind of input.
- AI workloads and text analysis covers the common workloads and the text techniques: key phrases, entities, sentiment and summarisation.
- Speech and computer vision capabilities covers speech recognition and synthesis, image analysis and image generation.
- Extracting information from content covers pulling structured data out of text, images, audio and video.
Ask what the input is and what the output should be. New content from a prompt is generative AI. Meaning pulled out of existing text is text analysis. Words from audio are speech recognition. Fields from a form are information extraction.
How to use these pages
Each topic page has the same shape, so you can skim for what you need.
- What it is: the idea in plain words.
- How it works: the mechanism, with examples and tables.
- What to remember for the exam: the facts to memorise and how questions are phrased.
- Common mistakes: the classic traps.
- Check yourself: three questions to test your understanding.
Ready for more practice?
The full Microsoft AI-901 question bank has 511 questions, with timed practice exams and an explanation for every answer.
This study guide is independent and is not affiliated with, endorsed by, or sponsored by Microsoft. Vendor names, certification names and exam codes are trademarks of their respective owners and are used for identification only.