The Integrated AI-900 (Microsoft Azure AI Fundamentals) Study Guide [120 Minute Read]

A First-Principles Approach to Artificial Intelligence on Azure

Welcome to 'The Integrated AI-900 Study Guide.' This guide moves beyond surface-level memorization. It is designed to build a robust mental model of artificial intelligence, understanding the 'why' behind every technology and design decision.

Rather than presenting disconnected facts, we begin with foundational principles that explain how AI actually works. Once you understand these core concepts, the specific Azure AI services and capabilities become logical extensions rather than arbitrary features to memorize.

Each topic is aligned with the official Microsoft AI-900 Exam Objectives, targeting the specific cognitive skills required for success.

Audience: Anyone interested in AI fundamentals. No data science or software engineering experience required. Basic familiarity with cloud concepts and client-server applications is helpful.

Passing Score: 700 out of 1000

Exam Domain Weights

💡 Study Strategy: The exam weights are relatively balanced, but Machine Learning and Generative AI concepts appear throughout multiple domains. Master the fundamentals in Phase 1 first—they unlock everything else.

---

(Table of Contents - For Reference)

• Phase 1: First Principles of Artificial Intelligence
  • 1.1. What Is Artificial Intelligence?
    • 1.1.1. The Core Abstraction: Pattern Recognition
    • 1.1.2. The AI Hierarchy: AI → ML → Deep Learning → Transformers
  • 1.2. The Input-Output Framework
    • 1.2.1. Input Modality Determines Workload Category
    • 1.2.2. Output Structure Determines Capability Type
  • 1.3. The Learning Paradigm
    • 1.3.1. Supervised Learning: Learning with Answers
    • 1.3.2. Unsupervised Learning: Finding Hidden Patterns
  • 1.4. The Azure AI Service Model
    • 1.4.1. Pre-built vs Custom Models
    • 1.4.2. The Universal Service Pattern
  • 1.5. Reflection Checkpoint: First Principles Mastery
• Phase 2: Describe Artificial Intelligence Workloads and Considerations (15-20%)
  • 2.1. AI Workload Categories
    • 2.1.1. Computer Vision Workloads
    • 2.1.2. Natural Language Processing Workloads
    • 2.1.3. Document Processing Workloads
    • 2.1.4. Generative AI Workloads
  • 2.2. Responsible AI Principles
    • 2.2.1. The Six Principles of Responsible AI
    • 2.2.2. Applying Principles to Real Scenarios
  • 2.3. Reflection Checkpoint: AI Workloads and Responsible AI Mastery
• Phase 3: Describe Fundamental Principles of Machine Learning on Azure (20-25%)
  • 3.1. Machine Learning Techniques
    • 3.1.1. Regression: Predicting Numbers
    • 3.1.2. Classification: Predicting Categories
    • 3.1.3. Clustering: Finding Natural Groups
    • 3.1.4. Deep Learning and Neural Networks
    • 3.1.5. The Transformer Architecture
  • 3.2. Core Machine Learning Concepts
    • 3.2.1. Features and Labels
    • 3.2.2. Training and Validation Datasets
  • 3.3. Azure Machine Learning Capabilities
    • 3.3.1. Automated Machine Learning
    • 3.3.2. Data and Compute Services
    • 3.3.3. Model Management and Deployment
  • 3.4. Reflection Checkpoint: Machine Learning Mastery
• Phase 4: Describe Features of Computer Vision Workloads on Azure (15-20%)
  • 4.1. Computer Vision Solution Types
    • 4.1.1. Image Classification
    • 4.1.2. Object Detection
    • 4.1.3. Optical Character Recognition (OCR)
    • 4.1.4. Facial Detection and Analysis
  • 4.2. Azure Computer Vision Services
    • 4.2.1. Azure AI Vision Service
    • 4.2.2. Azure AI Vision Advanced Features
    • 4.2.3. Azure AI Face Service
  • 4.3. Reflection Checkpoint: Computer Vision Mastery
• Phase 5: Describe Features of NLP Workloads on Azure (15-20%)
  • 5.1. NLP Workload Scenarios
    • 5.1.1. Key Phrase Extraction
    • 5.1.2. Entity Recognition
    • 5.1.3. Sentiment Analysis
    • 5.1.4. Language Modeling and Detection
    • 5.1.5. Speech Recognition and Synthesis
    • 5.1.6. Translation
  • 5.2. Azure NLP Services
    • 5.2.1. Azure AI Language Service
    • 5.2.2. Language Detection and Custom Features
    • 5.2.3. Question Answering and Text Analytics
    • 5.2.4. Azure AI Speech Service
  • 5.3. Reflection Checkpoint: NLP Mastery
• Phase 6: Describe Features of Generative AI Workloads on Azure (15-20%)
  • 6.1. Generative AI Fundamentals
    • 6.1.1. Features of Generative AI Models
    • 6.1.2. Model Parameters and Prompt Engineering
    • 6.1.3. Common Generative AI Scenarios
    • 6.1.4. Embeddings and Retrieval-Augmented Generation
    • 6.1.5. The Four-Layer Responsible AI Model for Generative AI
    • 6.1.6. Generative AI Risks and Mitigations
    • 6.1.7. Ethical and Governance Considerations
  • 6.2. Azure Generative AI Services
    • 6.2.1. Azure AI Foundry
    • 6.2.2. Azure OpenAI Service
    • 6.2.3. Azure AI Foundry Model Catalog
  • 6.3. Reflection Checkpoint: Generative AI Mastery
• Phase 7: Exam Readiness
  • 7.1. Exam Structure and Question Types
  • 7.2. Common Exam Patterns and Distractors
  • 7.3. Practice Questions: AI Concepts & Responsible AI
  • 7.4. Practice Questions: Machine Learning
  • 7.5. Practice Questions: Computer Vision
  • 7.6. Practice Questions: NLP
  • 7.7. Practice Questions: Generative AI
• Phase 8: Comprehensive Glossary

---