6.1 Domain Summary and Key Comparisons
Key Takeaways
- The AI-900 has five domains; after the May 2025 update the weights are AI workloads 15-20%, ML 15-20%, computer vision 15-20%, NLP 15-20%, and generative AI 20-25% (the heaviest).
- Map each task to its Azure service: Vision = images, Language = text, Speech = audio, Translator = text translation, OpenAI = generation, AI Search = retrieval, Document Intelligence = forms.
- ML types: regression predicts numbers, classification predicts categories (supervised), clustering discovers groups (unsupervised).
- Six responsible AI principles — fairness, reliability/safety, privacy/security, inclusiveness, transparency, accountability — are tested across ALL domains.
- Know the renames (LUIS->CLU, QnA Maker->Custom QA, Form Recognizer->Document Intelligence, Cognitive Search->AI Search, AI Studio->AI Foundry); the AI-900 retires June 30, 2026 and is replaced by AI-901, passing score 700/1000.
Last updated: June 2026
Quick Answer: This section consolidates the most important concepts, comparisons, and Azure service mappings from all five AI-900 domains into a single review. Use it as your final reference before the exam.
Exam Domain Weights (May 2025 update)
| Domain | Weight |
|---|---|
| Describe AI workloads and considerations | 15-20% |
| Describe fundamental principles of machine learning on Azure | 15-20% |
| Describe features of computer vision workloads on Azure | 15-20% |
| Describe features of NLP workloads on Azure | 15-20% |
| Describe features of generative AI workloads on Azure | 20-25% |
Generative AI is now the single heaviest domain (20-25%). Note that the AI-900 exam retires on June 30, 2026 and is replaced by AI-901 — pass before that date or plan for the successor exam. Passing score is 700/1000.
Domain 1: AI Workloads and Considerations (15-20%)
- AI is the broadest umbrella -> ML is a subset -> Deep Learning is a subset of ML -> Generative AI is a subset of DL.
- Common AI workloads: prediction/regression, classification, anomaly detection, computer vision, NLP, conversational AI, and generative AI.
- Six responsible AI principles: fairness, reliability/safety, privacy/security, inclusiveness, transparency, accountability.
- Responsible AI appears across ALL five domains, not just Domain 1; Azure AI Content Safety detects harmful content in four categories.
Domain 2: Machine Learning Principles (15-20%)
| Concept | Key Point |
|---|---|
| Features vs. Labels | Features = inputs; Labels = outputs (what you predict) |
| Training / Validation / Test | Training teaches, validation tunes, test evaluates |
| Overfitting | Great on training data, poor on new data |
| Supervised learning | Labeled data (regression, classification) |
| Unsupervised learning | Unlabeled data (clustering) |
ML Type Selection
| Question | Output Type | ML Type |
|---|---|---|
| "How much?" / "How many?" | Continuous number | Regression |
| "Which category?" / "Is it X?" | Discrete class | Classification |
| "What groups exist?" | Unknown groups | Clustering |
Evaluation Metrics
| ML Type | Metrics | Direction |
|---|---|---|
| Regression | MAE, RMSE, R-squared | MAE/RMSE lower; R-squared higher |
| Classification | Accuracy, Precision, Recall, F1 | Higher |
| Clustering | Silhouette score | Higher (closer to 1) |
Azure ML Features
| Feature | What It Does | Code Required |
|---|---|---|
| AutoML | Automatically selects the best algorithm | No |
| Designer | Drag-and-drop pipeline builder | No |
| Notebooks | Code-based ML development | Yes (Python) |
Domain 3: Computer Vision (15-20%)
| Task | Output | When to Use |
|---|---|---|
| Image classification | Single label | "What is this image?" |
| Object detection | Objects + bounding boxes | "Where are the objects?" |
| Semantic segmentation | Per-pixel classification | "What is every pixel?" |
| OCR | Extracted text | "What text is in this image?" |
| Facial detection | Face locations + attributes | "Where are the faces?" |
| Service | Purpose | Training Required |
|---|---|---|
| Azure AI Vision | Pre-built image analysis (caption, tag, detect, OCR) | No |
| Azure AI Custom Vision | Train custom classification/detection models | Yes (your images) |
| Azure AI Face | Face detection, verification, identification | No (gated access) |
| Azure AI Document Intelligence | Extract structured data from documents | Pre-built or custom |
Domain 4: NLP (15-20%)
| Task | What It Extracts | Example |
|---|---|---|
| Sentiment analysis | Positive/negative/neutral | "Great product!" -> Positive |
| Key phrase extraction | Main topics | Article -> ["AI", "Azure", "cloud"] |
| Named entity recognition | People, places, dates (classified) | "John in Paris on March 5" -> Person, Location, Date |
| Language detection | Which language | "Bonjour" -> French |
| PII detection | Sensitive personal data | SSNs, phone numbers, emails |
| Service | Purpose |
|---|---|
| Azure AI Language | Text analysis (sentiment, NER, CLU, Q&A, PII) |
| Azure AI Speech | Audio (STT, TTS, speech translation, speaker recognition) |
| Azure AI Translator | Text translation (100+ languages) |
Key Terminology Updates
| Old Name (Retired) | Current Name |
|---|---|
| LUIS | Conversational Language Understanding (CLU) |
| QnA Maker | Custom Question Answering |
| Form Recognizer | Document Intelligence |
| Cognitive Search | Azure AI Search |
| Cognitive Services | Azure AI Services |
| Azure AI Studio | Azure AI Foundry |
On the Exam: Microsoft has renamed many services. Know the CURRENT names; if a question uses an old name, recognize the modern equivalent.
Domain 5: Generative AI (20-25%)
| Concept | Key Point |
|---|---|
| LLMs | Large Language Models that predict the next token |
| Transformers | Architecture using self-attention for context |
| GPT | Generative Pre-trained Transformer |
| Hallucination | False but plausible output |
| RAG | Retrieve docs -> add to prompt -> generate grounded answer |
| Grounding | Provide factual data to reduce hallucinations |
| Prompt engineering | Craft prompts to control behavior |
| Content filters | Scan prompts AND responses (hate/violence/sexual/self-harm) |
| Temperature | 0 = deterministic; higher = creative |
| Embeddings | Vector representations of text for semantic search |
Prompt Engineering Techniques
| Technique | What It Does | When to Use |
|---|---|---|
| Zero-shot | No examples | Simple, well-defined tasks |
| Few-shot | Provide examples | Pattern/format control |
| Chain-of-thought | "Think step by step" | Math, logic, reasoning |
| Grounding | Provide context documents | Factual accuracy |
| System message | Set persona and constraints | Control behavior |
Master Comparison: All Azure AI Services
| Azure Service | Category | What It Does |
|---|---|---|
| Azure AI Vision | Vision | Pre-built image analysis, OCR, spatial analysis |
| Azure AI Custom Vision | Vision | Train custom image models |
| Azure AI Face | Vision | Face detection, verification, identification |
| Azure AI Language | NLP | Sentiment, NER, CLU, Q&A, PII |
| Azure AI Speech | NLP | STT, TTS, speech translation, speaker recognition |
| Azure AI Translator | NLP | Text translation (100+ languages) |
| Azure AI Document Intelligence | Document | Extract structured data from forms |
| Azure AI Search | Search | Vector/hybrid search, semantic ranking, RAG retrieval |
| Azure AI Content Safety | Safety | Detect harmful content in text/images |
| Azure OpenAI Service | Generative | GPT models, embeddings, image generation |
| Azure Machine Learning | ML | AutoML, Designer, training/deployment |
| Azure AI Foundry | Portal | Build, evaluate, deploy AI applications |
Test Your Knowledge
A company wants to analyze customer feedback emails to identify the main topics being discussed, determine if the feedback is positive or negative, and identify any people or products mentioned by name. Which THREE Azure AI Language features should they use?
A
B
C
D
Test Your Knowledge
Which of the following correctly describes the relationship between Azure AI Vision and Azure AI Custom Vision?
A
B
C
D
Test Your Knowledge
After the May 2025 skills update, which AI-900 domain carries the largest share of the exam?
A
B
C
D
Test Your Knowledge
MatchingMatch each machine learning type to its correct description:
Match each item on the left with the correct item on the right
1
Regression
2
Classification
3
Clustering