Complete AI-900 study guide — AI workloads, machine learning, computer vision, NLP, and generative AI on Azure.
This guide works best as a loop: read a chapter, test yourself with practice questions, look up unfamiliar terms in the glossary, then move to the next chapter.
100 chapters covering every exam objective. Each chapter includes key concepts, exam tips, common traps, comparison tables, and a 5-question quiz at the end.
Start Chapter 1Free timed and untimed practice with instant feedback and full explanations. Pick 10–120 questions per session. Filter by domain to drill your weak areas.
Go to practice testEvery AI-900 term defined and searchable. Use it when a chapter mentions a concept you haven't seen before or want a quick refresher on.
Browse glossaryExam blueprint, domain weights, passing score, duration, cost, and registration links. Start here if you're new to this certification.
View exam guide14 chapters
What is Artificial Intelligence?
Objective 1.1 · AI Workloads
Types of AI Workloads
Objective 1.1
Responsible AI Principles
Objective 1.2 · AI Workloads
AI vs Machine Learning vs Deep Learning
Objective 1.1
Narrow AI vs General AI
Objective 1.1
AI Use Cases: Prediction, Classification, Generation
Objective 1.1
AI Reliability, Safety, and Privacy
Objective 1.2
AI Transparency and Explainability
Objective 1.2
AI Accountability and Human Oversight
Objective 1.2
Azure AI Services Overview
Objective 1.1
Cognitive Services Pricing and Tiers
Objective 1.1
AI Infrastructure Costs and GPU Compute
Objective 1.1
AI-900 Exam Strategy and Scenario Questions
Objective 1.1
ML Ethics, Fairness, and Inclusiveness
Objective 1.2
28 chapters
Machine Learning Core Concepts
Objective 2.1 · Machine Learning
Supervised vs Unsupervised Learning
Objective 2.1
Regression and Classification
Objective 2.2 · Machine Learning
Deep Learning and Neural Networks
Objective 2.3 · Machine Learning
Azure Machine Learning Studio
Objective 2.4 · Machine Learning
Automated ML (AutoML)
Objective 2.4
Clustering
Objective 2.2
Features, Labels, and Training Data
Objective 2.1
Training, Validation, and Test Data Splits
Objective 2.1
Overfitting, Underfitting, and Model Complexity
Objective 2.1
ML Evaluation Metrics: Accuracy, Precision, Recall
Objective 2.2
Confusion Matrix and ROC Curve
Objective 2.2
Decision Trees and Random Forests
Objective 2.2
Linear vs Logistic Regression
Objective 2.2
Anomaly Detection
Objective 2.2
Recommendation Systems
Objective 2.2
Time Series Forecasting
Objective 2.2
Convolutional Neural Networks (CNN)
Objective 2.3
RNNs and Transformer Architecture
Objective 2.3
Transfer Learning and Pre-Trained Models
Objective 2.3
Azure Machine Learning Workspace
Objective 2.4
Azure ML Designer: Drag-and-Drop ML
Objective 2.4
Azure ML Notebooks and Compute Clusters
Objective 2.4
Azure ML Endpoints: Real-Time and Batch
Objective 2.4
Responsible AI Dashboard in Azure ML
Objective 2.4
No-Code AI Tools: Lobe, Teachable Machine
Objective 2.4
MLOps Concepts: Model Registry and Monitoring
Objective 2.4
Azure ML Pipelines for Batch Inference
Objective 2.4
17 chapters
What is Computer Vision?
Objective 3.1 · Computer Vision
Azure AI Vision Service
Objective 3.2 · Computer Vision
Azure Document Intelligence
Objective 3.4 · Computer Vision
Azure Custom Vision
Objective 3.5 · Computer Vision
Azure Face Service
Objective 3.3 · Computer Vision
Image Classification Tasks
Objective 3.1
Object Detection and Bounding Boxes
Objective 3.1
Semantic Segmentation
Objective 3.1
Optical Character Recognition (OCR)
Objective 3.2
Image Analysis: Tags, Captions, Objects
Objective 3.2
Spatial Analysis and Video Insights
Objective 3.2
Face Detection vs Face Recognition
Objective 3.3
Face Attributes and Emotion Detection
Objective 3.3
Document Layout Analysis
Objective 3.4
Pre-Built Models: Invoices, Receipts, IDs
Objective 3.4
Custom Vision Models: Training and Evaluation
Objective 3.5
Custom Image Classification vs Object Detection
Objective 3.5
21 chapters
What is NLP?
Objective 4.1 · NLP
Azure AI Language Service
Objective 4.2 · NLP
Sentiment Analysis and Key Phrase Extraction
Objective 4.3 · NLP
Azure AI Speech Service
Objective 4.4 · NLP
Azure AI Translator
Objective 4.5 · NLP
Azure Bot Service and QnA Maker
Objective 4.6 · NLP
Tokenization and Text Normalization
Objective 4.1
Named Entity Recognition (NER)
Objective 4.2
Entity Linking and Knowledge Base
Objective 4.2
Question Answering with Language Service
Objective 4.3
Conversational Language Understanding (CLU)
Objective 4.6
Speech to Text and Custom Speech
Objective 4.4
Text to Speech and Neural Voices
Objective 4.4
Speech Translation
Objective 4.5
Document Translation
Objective 4.5
Custom Translator
Objective 4.5
LUIS vs Conversational Language Understanding
Objective 4.6
Power Virtual Agents and Azure Bot Framework
Objective 4.6
Language Detection
Objective 4.2
PII Detection and Extraction
Objective 4.3
Extractive vs Abstractive Text Summarisation
Objective 4.3
20 chapters
What is Generative AI?
Objective 5.1 · Generative AI
Large Language Models (LLMs)
Objective 5.1
Azure OpenAI Service
Objective 5.2 · Generative AI
Prompt Engineering Fundamentals
Objective 5.3 · Generative AI
Microsoft Copilot and AI-900
Objective 5.4 · Generative AI
Types of Generative AI: Text, Image, Code, Audio
Objective 5.1
Foundation Models and Fine-Tuning
Objective 5.1
GPT Models: GPT-3.5, GPT-4, o-Series
Objective 5.2
DALL-E for Image Generation
Objective 5.2
Azure OpenAI Deployments and API Access
Objective 5.2
Zero-Shot, Few-Shot, and Chain-of-Thought Prompts
Objective 5.3
System Messages and Grounding Prompts
Objective 5.3
Retrieval Augmented Generation (RAG)
Objective 5.3
Embeddings and Vector Search
Objective 5.3
Azure AI Content Safety
Objective 5.4
Microsoft Copilot Ecosystem
Objective 5.4
Microsoft Copilot Studio
Objective 5.4
Azure AI Foundry (Azure AI Hub)
Objective 5.2
Azure AI Search for RAG Applications
Objective 5.3
Azure OpenAI Fine-Tuning
Objective 5.2
Free AI-900 practice questions with full explanations. Test what you learn chapter by chapter.
AI-900 Practice Questions