Applied Azure AI: Building Intelligent Solutions with Foundry Tools
Author: Dr. Frazier A. Smith, Central Piedmont Community College
About This Book
Applied Azure AI: Building Intelligent Solutions with Foundry Tools is an Open Educational Resource (OER) textbook designed for students pursuing practical competency in Microsoft Azure’s artificial intelligence platform. The book is aligned to the Microsoft AI-103 (Azure AI App and Agent Developer Associate) certification and covers the breadth of Azure AI development, from generative AI foundations and retrieval-augmented generation to AI agents, multi-agent orchestration, speech, computer vision, and multimodal information extraction.
This textbook takes a hands-on, applied approach. Rather than treating AI concepts in the abstract, every chapter grounds its content in real-world scenarios and connects directly to Skillable Cloud Slice lab environments where students build, deploy, and evaluate Azure AI solutions. The goal is not simply to prepare students for a certification exam, but to develop the judgment and technical skill needed to architect intelligent solutions in professional settings.
The book is structured as eight chapters spanning eight weeks of instruction, making it suitable for a semester course or an accelerated professional development program. Each chapter builds on the previous one, beginning with model deployment and evaluation and progressing through increasingly sophisticated patterns — retrieval-augmented generation, responsible AI, AI agent development, multi-agent orchestration, text analysis, speech, vision, and multimodal information extraction — culminating in a capstone project that integrates multiple Foundry Tools and Azure AI capabilities.
How to Use This Book
This textbook follows a 1:1 chapter-to-week mapping. Students read Chapter 1 during Week 1, Chapter 2 during Week 2, and so on through all eight weeks of the course. Each chapter is designed to support that week’s complete learning cycle: readings introduce concepts and architecture, hands-on connection sections bridge to that week’s Skillable labs, and key takeaways reinforce the material assessed in that week’s quiz and assignments.
Each chapter follows a consistent structure: an introduction with a real-world scenario, learning objectives tied to module-level outcomes, core content sections with technical depth, a hands-on connections section that prepares students for lab work, key takeaways, and review questions. This predictable format helps students navigate the material efficiently and know what to expect each week.
Table of Contents
| Chapter | Description |
|---|---|
Microsoft Foundry model catalog, serverless and provisioned deployments, chat completion API configuration, model evaluation, and secure project setup fundamentals. |
|
Retrieval-augmented generation on your own data, document chunking and embedding strategies, grounding model responses in source content, content safety filters, and responsible-AI governance principles. |
|
The agent reasoning loop, building agents in the Microsoft Foundry portal and VS Code, defining and registering custom function tools, and managing the tool execution lifecycle. |
|
Model Context Protocol (MCP) tool connections, agentic retrieval with Foundry IQ knowledge bases, and integrating external knowledge sources into agent workflows. |
|
Microsoft Agent Framework abstractions, multi-agent orchestration patterns (sequential, parallel, orchestrator), and scoping a capstone proposal that integrates multiple services. |
|
Generative text analysis, building a text-analysis agent, Azure AI Speech (speech-to-text, text-to-speech), and building real-time conversational voice agents with Voice Live. |
|
Vision-enabled chat completions, image and video generation and editing, multimodal understanding, and selecting the right vision service for a given task. |
|
Content Understanding multimodal extraction, grounded representations (markdown/JSON) feeding RAG and Foundry IQ, multi-service architecture design, and the agentic-multimodal capstone. |