Skip to main content

Coverage Matrix

This matrix maps every official AI-900 exam skill to the challenge(s) that cover it. Use this to identify gaps in your preparation and target specific challenges for review.

Domain 1: Describe AI workloads and considerations (15–20%)

Identify features of common AI workloads

SkillChallenge(s)
Identify features of anomaly detection workloadsCh 01
Identify computer vision workloadsCh 01, Ch 03
Identify natural language processing workloadsCh 01, Ch 03
Identify knowledge mining workloadsCh 01, Ch 03
Identify generative AI workloadsCh 01, Ch 03
Identify document intelligence workloadsCh 01, Ch 03

Identify guiding principles for responsible AI

SkillChallenge(s)
Describe fairness considerationsCh 02
Describe reliability and safety considerationsCh 02
Describe privacy and security considerationsCh 02
Describe inclusiveness considerationsCh 02
Describe transparency considerationsCh 02
Describe accountability considerationsCh 02

Domain 2: Describe fundamental principles of machine learning on Azure (15–20%)

Describe fundamental principles of machine learning

SkillChallenge(s)
Identify features and labels in a datasetCh 05
Describe how training and validation datasets are usedCh 05
Describe how machine learning algorithms are used for model trainingCh 05
Describe model inference and deploymentCh 05

Describe Azure Machine Learning capabilities

SkillChallenge(s)
Describe capabilities of Azure Machine LearningCh 09
Describe data and compute resources in Azure MLCh 09
Describe jobs and pipelines in Azure MLCh 09

Machine learning model types

SkillChallenge(s)
Describe regression models — prediction of continuous valuesCh 05
Describe classification models — binary and multi-classCh 06
Describe clustering models — unsupervised groupingCh 07
Describe deep learning conceptsCh 08

Domain 3: Describe features of computer vision workloads on Azure (15–20%)

Identify types of computer vision solutions

SkillChallenge(s)
Identify types of computer vision solutionsCh 10, Ch 11
Describe Azure AI Vision service featuresCh 10

Image analysis

SkillChallenge(s)
Describe image classificationCh 10
Describe object detectionCh 11
Describe optical character recognition (OCR)Ch 12
Describe facial detection and analysisCh 13

Domain 4: Describe features of Natural Language Processing (NLP) workloads on Azure (15–20%)

Describe NLP capabilities

SkillChallenge(s)
Describe key phrase extractionCh 14
Describe entity recognitionCh 14
Describe sentiment analysisCh 15
Describe language modelingCh 15
Describe speech recognition and synthesisCh 16
Describe translation capabilitiesCh 17

Azure AI Language and Speech services

SkillChallenge(s)
Describe capabilities of Azure AI LanguageCh 18
Describe capabilities of Azure AI SpeechCh 16, Ch 18

Domain 5: Describe features of generative AI workloads on Azure (20–25%)

Describe generative AI concepts

SkillChallenge(s)
Describe generative AI fundamentalsCh 19
Identify features of large language models (LLMs)Ch 19
Describe Azure OpenAI ServiceCh 20
Describe Azure AI FoundryCh 21
Describe prompt engineering conceptsCh 22
Describe responsible generative AI considerationsCh 23

Summary: Challenge coverage

ChallengePrimary domainKey topics
Ch 01Domain 1AI workloads overview
Ch 02Domain 1Responsible AI principles
Ch 03Domain 1AI workloads deep-dive
Ch 04Domain 1Azure AI Services Overview
Ch 05Domain 2ML fundamentals, regression
Ch 06Domain 2Classification
Ch 07Domain 2Clustering
Ch 08Domain 2Deep learning
Ch 09Domain 2Azure Machine Learning
Ch 10Domain 3Image analysis, classification
Ch 11Domain 3Object detection
Ch 12Domain 3OCR
Ch 13Domain 3Face detection
Ch 14Domain 4Key phrase extraction, entity recognition
Ch 15Domain 4Sentiment analysis, language modeling
Ch 16Domain 4Speech recognition and synthesis
Ch 17Domain 4Translation
Ch 18Domain 4Azure AI Language & Speech services
Ch 19Domain 5Generative AI fundamentals
Ch 20Domain 5Azure OpenAI Service
Ch 21Domain 5Azure AI Foundry
Ch 22Domain 5Prompt engineering
Ch 23Domain 5Responsible generative AI
Ch 24AllCapstone challenge — full exam simulation
tip

Use this matrix to identify weak areas. If a domain feels unfamiliar, work through its challenges in sequence. If you're confident in a domain, jump straight to the later challenges for that section to validate your knowledge.