A comprehensive preparation repository for the Microsoft Certified: Azure AI Engineer Associate exam (AI-102). Built as a structured, hands-on study guide covering all official exam domains as of December 23, 2025.
The Big Shift: Azure AI has rebranded under Microsoft Foundry (formerly Azure AI Foundry / Azure AI Studio). The exam now heavily emphasizes Generative AI, RAG patterns, Agentic solutions, and Responsible AI governance.
No Azure Lab? This repo is designed so you can fully study and practice without an Azure subscription. All demo scripts run in simulation mode. Both full mock exams (100 questions total) require no Azure resources.
| Detail | Info |
|---|---|
| Exam Code | AI-102 |
| Certification | Microsoft Certified: Azure AI Engineer Associate |
| Duration | 100 minutes |
| Passing Score | 700 / 1000 |
| Price | ~$165 USD (varies by region) |
| Languages | Python or C# (your choice for code questions) |
| Renewal | Annually (free online assessment) |
| Domain | Weight | Study Guide |
|---|---|---|
| 1. Plan and Manage an Azure AI Solution | 20–25% | docs/domain-01-plan-manage.md |
| 2. Implement Generative AI Solutions | 15–20% | docs/domain-02-generative-ai.md |
| 3. Implement an Agentic Solution | 5–10% | docs/domain-03-agentic-solutions.md |
| 4. Implement Computer Vision Solutions | 10–15% | docs/domain-04-computer-vision.md |
| 5. Implement Natural Language Processing Solutions | 15–20% | docs/domain-05-nlp.md |
| 6. Implement Knowledge Mining & Information Extraction | 15–20% | docs/domain-06-knowledge-mining.md |
All learning resources in this repo work without an Azure subscription:
| Resource | No Azure Needed | With Azure |
|---|---|---|
| Domain study guides (all 6) | ✅ | ✅ |
| Mock Exam 1 (50 questions + answers) | ✅ | ✅ |
| Mock Exam 2 (50 questions + answers) | ✅ | ✅ |
| Demo scripts (simulation mode) | ✅ | ✅ |
| Quick Reference cheat sheet | ✅ | ✅ |
| Flashcards, quiz prompts, scenarios | ✅ | ✅ |
| Glossary and domain weights guide | ✅ | ✅ |
| Run demos with real API calls | ❌ | ✅ |
AI-102_Cert_Prep/
│
├── README.md ← Start here
├── QUICK-REFERENCE.md ← Print this before exam day
├── COURSE-PLAN.md ← 8-week structured study plan
├── CONTRIBUTING.md
├── LICENSE
│
├── docs/ ← Complete domain study guides
│ ├── domain-01-plan-manage.md ← 20-25% of exam
│ ├── domain-02-generative-ai.md ← 15-20%
│ ├── domain-03-agentic-solutions.md ← 5-10%
│ ├── domain-04-computer-vision.md ← 10-15%
│ ├── domain-05-nlp.md ← 15-20%
│ └── domain-06-knowledge-mining.md ← 15-20%
│
├── mocks/ ← NO AZURE NEEDED — Pure markdown exams
│ ├── full-mock-exam-01.md ← 50 questions, timed (100 min)
│ ├── full-mock-exam-02.md ← 50 questions, harder scenarios
│ └── answer-keys/
│ ├── mock-exam-01-answers.md ← Full explanations for all 50 answers
│ └── mock-exam-02-answers.md
│
├── demos/ ← SDK code demos (all run without Azure)
│ ├── 01-plan-manage/
│ │ ├── README.md
│ │ └── demo_plan_manage.py ← Language + Content Safety
│ ├── 02-generative-ai/
│ │ ├── README.md
│ │ ├── openai/demo_openai.py ← GPT-4o, DALL-E, parameters
│ │ └── rag/demo_rag.py ← Full RAG pipeline simulation
│ ├── 03-agentic-solutions/
│ │ ├── README.md
│ │ └── demo_agents.py ← Agent reasoning + tool calling
│ ├── 04-computer-vision/
│ │ ├── README.md
│ │ └── demo_vision.py ← Image analysis, Custom Vision
│ ├── 05-nlp/
│ │ ├── README.md
│ │ ├── language/demo_language.py ← Sentiment, NER, PII, language detection
│ │ ├── speech/demo_speech.py ← STT, TTS, SSML, translation
│ │ └── clu/demo_clu.py ← CLU intents, entities, CQA
│ └── 06-knowledge-mining/
│ ├── README.md
│ ├── demo_search.py ← AI Search pipeline + query types
│ └── demo_doc_intelligence.py ← Document Intelligence models
│
├── exam-metadata/
│ ├── ai-102-exam-objectives.md ← Official objectives (Dec 2025)
│ ├── domain-weights.md ← Study time allocation guide
│ └── key-terms-glossary.md ← 70+ term definitions
│
├── resources/
│ └── service-selection-guide.md ← "Which service for X?" scenarios
│
├── scripts/
│ └── setup-azure-resources.sh ← Optional: create Azure resources via CLI
│
├── src/
│ └── python/
│ ├── requirements.txt ← All Python dependencies
│ └── env-template.txt ← Environment variable template
│
└── .github/
├── SECURITY.md
└── prompts/ ← AI-powered study prompts
├── quiz-generator.md ← Generate practice questions with AI
├── scenario-builder.md ← Architecture scenario walkthroughs
└── flashcard-generator.md ← Active recall Q&A pairs
git clone https://github.com/YOUR_USERNAME/AI-102_Cert_Prep.git
cd AI-102_Cert_Prep# Open the week-by-week curriculum
open COURSE-PLAN.md# Start with the highest-weight domain
open docs/domain-01-plan-manage.mdcd demos/01-plan-manage
python demo_plan_manage.py
# Shows simulated output — no Azure credentials needed# Set a 100-minute timer, then open:
open mocks/full-mock-exam-01.md
# Check your answers:
open mocks/answer-keys/mock-exam-01-answers.md# Install dependencies
pip install -r src/python/requirements.txt
# Copy and fill in your credentials
cp src/python/env-template.txt .env
# Edit .env with your endpoints
# Run demos with real Azure APIs
python demos/01-plan-manage/demo_plan_manage.py| Resource | Purpose |
|---|---|
| QUICK-REFERENCE.md | Print this the night before exam |
| COURSE-PLAN.md | 8-week structured curriculum |
| mocks/full-mock-exam-01.md | 50-question practice exam |
| mocks/full-mock-exam-02.md | 50-question harder practice exam |
| resources/service-selection-guide.md | "Which service?" decision tables |
| exam-metadata/key-terms-glossary.md | 70+ term definitions |
| exam-metadata/domain-weights.md | Study time allocation by domain |
| .github/prompts/ | Generate more questions with AI |
| Service | Category | Exam Weight |
|---|---|---|
| Azure OpenAI (in Foundry Models) | Generative AI | ⭐⭐⭐⭐⭐ |
| Microsoft Foundry (AI Foundry) | Platform | ⭐⭐⭐⭐⭐ |
| Azure AI Language | NLP | ⭐⭐⭐⭐ |
| Azure AI Search | Knowledge Mining | ⭐⭐⭐⭐ |
| Azure Document Intelligence | Information Extraction | ⭐⭐⭐⭐ |
| Azure AI Vision | Computer Vision | ⭐⭐⭐ |
| Azure AI Speech | NLP / Speech | ⭐⭐⭐ |
| Azure AI Translator | NLP | ⭐⭐⭐ |
| Azure AI Content Safety | Responsible AI | ⭐⭐⭐ |
| Azure AI Video Indexer | Computer Vision | ⭐⭐ |
- Microsoft Foundry is the new name — exam uses both "Azure AI Foundry" and "Microsoft Foundry"
- Agentic solutions are now a standalone domain (5–10%) — don't skip it
- RAG appears heavily throughout Domains 2 & 6
- Responsible AI is embedded across ALL domains
- Code questions — choose Python OR C# at start; you cannot switch mid-exam
- DALL-E 3 only supports
n=1(one image per request) - Semantic search requires Standard S1 tier or higher
- Container deployment always requires
Eula=accept+ billing endpoint
| Old / Deprecated | Current Name |
|---|---|
| Azure AI Studio | Microsoft Foundry / Azure AI Foundry |
| Form Recognizer | Azure Document Intelligence |
| LUIS | CLU (Conversational Language Understanding) |
| QnA Maker | Custom Question Answering |
| Azure Cognitive Search | Azure AI Search |
| Azure Cognitive Services | Azure AI Services |
| Computer Vision API v3 | Azure Vision in Foundry Tools |
| Resource | Link |
|---|---|
| Exam Page | Microsoft Learn – AI-102 |
| Official Study Guide | AI-102 Study Guide |
| Free Practice Assessment | Practice Assessment |
| Azure AI Services Docs | Azure AI Services |
| Microsoft Foundry Docs | AI Foundry |
MIT License. See LICENSE for details.
Last updated: March 2026 · Exam objectives as of December 23, 2025