Azure AI Engineer Associate (AI-102) Practice Plan¶

Overview¶

Exam: AI-102 - Microsoft Azure AI Engineer Associate Timeline: 10-12 weeks Study Time: 10-15 hours per week Difficulty: Associate level Prerequisites: Programming experience (C# or Python), Azure fundamentals

Study Resources¶

Official Microsoft Learn AI-102 Path (FREE)
Azure AI Services Documentation
Complete Azure Study Resources
README.md in this directory

Week-by-Week Breakdown¶

Week 1-2: Plan and Manage Azure AI Solutions¶

Learning Objectives: - Select appropriate AI services - Configure AI service security - Create and configure AI services - Manage costs and monitor services

Hands-on Labs: - [ ] Create Azure AI Services multi-service resource - [ ] Configure authentication and authorization - [ ] Implement managed identities - [ ] Set up Key Vault for keys - [ ] Configure diagnostic logging - [ ] Monitor with Application Insights - [ ] Implement cost management alerts

Practice: 30-40 questions on planning and management

Week 3-4: Implement Computer Vision Solutions¶

Learning Objectives: - Analyze images with Computer Vision - Implement custom computer vision models - Analyze videos with Video Indexer - Detect faces and facial features

Hands-on Labs: - [ ] Analyze images using Computer Vision API - [ ] Extract text with OCR (Read API) - [ ] Train custom image classification model - [ ] Build object detection model - [ ] Detect and analyze faces - [ ] Use Video Indexer to analyze videos - [ ] Extract insights and metadata from videos

Practice: 50-60 questions on computer vision

Week 5-6: Implement Natural Language Processing¶

Learning Objectives: - Analyze text with Azure AI Language - Process speech with Azure AI Speech - Translate languages - Create question answering solutions

Hands-on Labs: - [ ] Extract key phrases and entities - [ ] Perform sentiment analysis - [ ] Detect language and PII - [ ] Implement speech-to-text - [ ] Build text-to-speech with SSML - [ ] Create custom speech models - [ ] Implement translation (text and speech) - [ ] Build question answering knowledge base

Practice: 60-70 questions on NLP

Week 7-8: Language Understanding and Conversational AI¶

Learning Objectives: - Create Language Understanding models - Implement intent recognition - Build conversational AI solutions - Optimize language models

Hands-on Labs: - [ ] Create intents and add utterances - [ ] Define entities (prebuilt and custom) - [ ] Train and test Language Understanding model - [ ] Publish and consume from client app - [ ] Optimize model performance - [ ] Implement conversation flow - [ ] Add multi-turn conversations - [ ] Export and version models

Practice: 50-60 questions on language understanding

Week 9: Knowledge Mining and Document Intelligence¶

Learning Objectives: - Implement Azure Cognitive Search - Create enrichment pipelines - Implement Document Intelligence

Hands-on Labs: - [ ] Create search service and index - [ ] Configure data sources and indexers - [ ] Implement skillsets with built-in skills - [ ] Create custom skills - [ ] Query search index - [ ] Use prebuilt Document Intelligence models - [ ] Train custom document models - [ ] Create composed models

Practice: 40-50 questions on knowledge mining

Week 10: Implement Generative AI Solutions¶

Learning Objectives: - Use Azure OpenAI Service - Generate content with GPT models - Generate images with DALL-E - Optimize generative AI solutions

Hands-on Labs: - [ ] Provision Azure OpenAI resource - [ ] Deploy GPT-3.5 and GPT-4 models - [ ] Generate natural language responses - [ ] Implement code generation - [ ] Generate images with DALL-E - [ ] Apply prompt engineering techniques - [ ] Configure parameters (temperature, top_p) - [ ] Implement RAG with your own data

Practice: 40-50 questions on generative AI

Week 11-12: Final Practice and Review¶

Focus Areas: - [ ] Complete 4-5 full practice exams - [ ] Build end-to-end AI solution project - [ ] Review all weak areas - [ ] Practice SDK coding scenarios - [ ] Review responsible AI principles

Practice Exams: - [ ] Exam 1: _% - [ ] Exam 2: % - [ ] Exam 3: __% - [ ] Exam 4: _____%

Key Topics Checklist¶

Plan and Manage AI Solutions (15-20%)¶

Select appropriate AI service for each scenario
Plan AI service security requirements
Configure authentication (keys, Azure AD)
Implement managed identities
Configure diagnostic logging
Monitor AI service performance
Manage costs

Computer Vision Solutions (20-25%)¶

Natural Language Processing (30-35%)¶

Text analysis (entities, key phrases, sentiment)
Language detection and PII detection
Speech-to-text and text-to-speech
Custom speech models
Language translation (text and speech)
Create and manage Language Understanding models
Build question answering solutions
Multi-turn conversations

Knowledge Mining (10-15%)¶

Create and configure Azure Cognitive Search
Define indexes and indexers
Implement skillsets
Create custom skills
Query search indexes
Use Document Intelligence prebuilt models
Train custom document models

Generative AI (10-15%)¶

Critical Skills¶

Programming Skills¶

Proficiency in C# or Python
REST API consumption
SDK usage (Azure.AI.* libraries)
Async programming patterns
Error handling and retries

AI Concepts¶

Understand confidence scores
Model training and evaluation
Precision vs recall
Overfitting and underfitting
Prompt engineering techniques

Azure Integration¶

Managed identity implementation
Key Vault integration
Application Insights monitoring
Secure endpoint configuration

Hands-on Project Ideas¶

Intelligent Document Processing: Extract data from forms using Document Intelligence
Customer Service Bot: QnA Maker + Language Understanding + Speech
Content Moderation System: Computer Vision + Text Analytics
Video Analytics Platform: Video Indexer + custom skills
Enterprise Search: Cognitive Search with custom skillsets

Exam Tips¶

Know when to use each AI service
Understand SDK code patterns
Practice with both REST API and SDKs
Expect code-focused questions
Understand responsible AI principles
Know service limits and quotas

Additional Resources¶

Success Strategy: Build real AI applications using Azure AI services. Focus on SDK programming and service integration patterns.