AI-102 Exam Study Guide

---

DISCLAIMER

This study guide was generated with assistance from GitHub Copilot.

While this guide references official Microsoft Azure documentation, validate details against the source links included. This is a supplementary study resource—API versions, quotas, endpoints, and pricing can change. Always check the official docs.

---

Split-page version: each numbered section is now its own page under the guide/ folder.

• Use the Table of Contents below to navigate.
• Each page has Previous / Next / Table of Contents links.
• Full single-page backup: AI-102_Study_guide_full.md

Table of Contents

• AI-102 Exam Study Guide
  • DISCLAIMER
  • Table of Contents
  • 1. File Types and Size Limitations
    • Document Intelligence (Form Recognizer)
    • Custom Vision
    • Language Services (Text Analytics)
    • Speech Services
    • Azure AI Search
    • Azure Storage (for AI services)
    • Azure Video Indexer (Video processing)
  • 2. Control Plane - REST API Calls
    • API key note
    • Common REST Endpoint structure
    • Document Intelligence — Analyze (REST example)
    • Language Services — AnalyzeTextAsync (overview & REST)
    • Custom Vision — Prediction (REST)
    • Speech-to-Text (REST)
    • Azure OpenAI — Chat completions (REST)
  • 3. Content Filtering and Message Flagging
  • 4. Multi-Language Support for Audio and Video
  • 5. Model Selection Criteria (Simplified)
  • 6. Azure Security: Key Rotation
  • 7. Synonyms in Document Intelligence
  • 8. Confidence Scoring (Short)
  • 9. Azure AI Language (Question Answering)
  • 10. Token Calculation & Max Token Behavior
  • 11. Image Model Training (Vision)
  • 12. Text / Image Processing Methods (REST)
  • 13. Semantic Kernel — Prompt Template Formats
  • 14. SSML for Speech — Styles & Example
    • SSML attribute quick reference (Azure Speech)
  • 15. AI Search Indexing (Summary)
    • What resource types can be queried with Azure AI Search?
    • Leverage auto tag and classification in vision search
  • Video (Azure Video Indexer & video processing)
    • When to run Video Indexing (and what you get)
  • 16. Quick Reference Tables
    • API Key Quick Reference
    • File Size / Limits Summary
    • Video File Outputs (Video Indexer)
    • Model Selection (cheat sheet)
    • AnalyzeTextAsync() Snapshot
    • Token Behavior Summary
    • Confidence Score Thresholds (guidance)
    • JSONPath Wildcard Examples
  • 17. Entity Linking (Wikipedia links from text)
  • 18. Import/Export Azure AI Language Projects
  • 19. Foundry: Scope model access for multiple APIs
  • Useful Links (official docs)

---

🎯 Interactive Quiz

This repository includes an interactive quiz agent to help you test your knowledge and prepare for the AI-102 certification exam.

How to Use the Quiz Agent

1. In VS Code with GitHub Copilot Chat: Open the chat panel and invoke the quiz agent by typing:

   @quiz start a quiz
   

2. Quiz Features:

   • Multiple-choice questions covering all study guide topics
   • Detailed explanations for correct and incorrect answers
   • Progress tracking and score reporting
   • Links to official Microsoft Learn documentation

MCP Server Configuration

The quiz agent uses the Microsoft Learn MCP (Model Context Protocol) server to fetch the latest official documentation. The configuration is included in .vscode/mcp.json.

To enable the MCP server:

1. Ensure you have VS Code with GitHub Copilot Chat extension installed
2. The MCP configuration will be automatically loaded from .vscode/mcp.json
3. The server provides access to:
   • microsoft_docs_search - Search Microsoft Learn documentation
   • microsoft_docs_fetch - Fetch complete documentation pages
   • microsoft_code_sample_search - Find official code samples

Manual MCP Configuration (if needed):

Add to your VS Code settings or .vscode/mcp.json:

{
  "servers": {
    "microsoft-docs": {
      "type": "http",
      "url": "https://mcp.microsoft.com/sse"
    }
  }
}

---