Azure AI Engineer Associate (AI-102) Study Guide In-Depth Certification Guide and Practice (Renaldi Gondosubroto)（Z-Library）

Page 1

Renaldi Gondosubroto Azure AI Engineer Associate (AI-102) Study Guide In-Depth Certif ication Guide and Practice

Page 2

9 7 8 1 0 9 8 1 6 9 2 6 8 5 5 9 9 9 ISBN: 978-1-098-16926-8 US $59.99 CAN $74.99 CLOUD COMPUTING With the GenAI boom showing no sign of letup, the demand for AI skills will only increase with time and innovation. Microsoft Azure leads the pack with services for developing and deploying AI solutions, so professionals looking to break into this field should consider pursuing certification as an Azure AI Engineer Associate. Azure’s AI-102 exam isn’t a piece of cake, but author Renaldi Gondosubroto makes it a great deal more approachable with this comprehensive study guide. Packed with expert guidance, it covers everything you’ll need to know to pass the exam. You’ll dive deep into all the phases of AI solutions development, from requirements definition and design to development, deployment, and integration, along with maintenance, performance tuning, and monitoring throughout. The book also takes you through practical implementation of these systems, covering decision support, computer vision, natural language processing, knowledge mining, document intelligence, and generative AI solutions. • Understand the core concepts of Azure AI services • Develop and deploy AI solutions within Azure’s environment • Explore integration and security practices with Azure AI services • Optimize and troubleshoot AI models on Azure • Gain knowledge about building GenAI solutions on Azure and put it into practice Azure AI Engineer Associate (AI-102) Study Guide “ Cuts through the ‘AI gold rush’ hype to deliver practical, real-world advice. This study guide ef fectively translates Azure’s powerful AI toolkit into tangible business value, preparing engineers not just for the AI-102 exam but for the actual demands of the f ield.” Prashanth Josyula, principal member of technical staff, Salesforce “This guide is a comprehensive, technically sound, and hands-on companion for anyone preparing for the AI-102 exam. It not only demystif ies Azure AI services with clarity and practical exercises but also ensures readers understand real-world use cases and responsible AI principles. Renaldi has done an excellent job blending certif ication objectives with applied knowledge—making this book an essential resource for aspiring Azure AI engineers.” Shashank Pawar, director of data and analytics, Microsoft Renaldi Gondosubroto is an accomplished software engineer and developer advocate with more than a decade of experience developing AI solutions. He is a Microsoft Certified Trainer and holds a master of science in computer science from Columbia University and all 20 certifications in Microsoft Azure.

Page 3

Renaldi Gondosubroto Azure AI Engineer Associate (AI-102) Study Guide In-Depth Certification Guide and Practice

Page 4

978-1-098-16926-8 [LSI] Azure AI Engineer Associate (AI-102) Study Guide by Renaldi Gondosubroto Copyright © 2025 Renaldi Gondosubroto. All rights reserved. Published by O’Reilly Media, Inc., 141 Stony Circle, Suite 195, Santa Rosa, CA 95401. O’Reilly books may be purchased for educational, business, or sales promotional use. Online editions are also available for most titles (https://oreilly.com). For more information, contact our corporate/institu‐ tional sales department: 800-998-9938 or corporate@oreilly.com. Acquisitions Editor: Megan Laddusaw Development Editor: Angela Rufino Production Editor: Ashley Stussy Copyeditor: Doug McNair Proofreader: Rachel Wheeler Indexer: Judith McConville Cover Designer: Susan Brown Cover Illustrator: José Marzan Jr. Interior Designer: David Futato Interior Illustrator: Kate Dullea September 2025: First Edition Revision History for the First Edition 2025-09-09: First Release See https://oreilly.com/catalog/errata.csp?isbn=9781098169268 for release details. The O’Reilly logo is a registered trademark of O’Reilly Media, Inc. Azure AI Engineer Associate (AI-102) Study Guide, the cover image, and related trade dress are trademarks of O’Reilly Media, Inc. The views expressed in this work are those of the author and do not represent the publisher’s views. While the publisher and the author have used good faith efforts to ensure that the information and instructions contained in this work are accurate, the publisher and the author disclaim all responsibility for errors or omissions, including without limitation responsibility for damages resulting from the use of or reliance on this work. Use of the information and instructions contained in this work is at your own risk. If any code samples or other technology this work contains or describes is subject to open source licenses or the intellectual property rights of others, it is your responsibility to ensure that your use thereof complies with such licenses and/or rights.

Page 5

Table of Contents Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix 1. Introduction to AI Solutions on Microsoft Azure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Introduction to the AI Landscape 2 The Minimally Qualified Candidate for the AI-102 Exam 3 Where We Are Now with AI 4 Fundamental AI Concepts 5 The Six Core Principles: Considerations for Developing AI Responsibly 9 Explainable AI Techniques 11 Tech Setup 12 Setting Up Your Microsoft Azure Account 12 Configuring Your Azure AI Environment 13 Configuring Your Local Development Environment 16 Gaining an Understanding of Azure AI’s Capabilities 18 The Capabilities of Microsoft Azure’s AI Services 19 Consuming AI Services 24 Authentication and Security 25 Billing and Cost Management 27 Your Responsibilities as an AI Engineer 27 Meeting Challenges and Managing Risks 28 Continuous Learning and Collaboration 29 User-Centered Design 29 Practical: Running a Text Analytics Service 30 Chapter Review 34 Chapter Quiz 35 iii

Page 6

2. Planning and Managing AI Solutions in Microsoft Azure. . . . . . . . . . . . . . . . . . . . . . . . . . 37 The Azure AI Project Lifecycle 38 Requirements definition and design 39 Development 40 Deployment 40 Integration 41 Maintenance 41 Performance Tuning 41 Monitoring 42 Practical: Designing an AI Solution 42 A Simple Example of Solution Design 44 Weighing the Trade-offs 45 Planning and Configuring Access and Security 45 Implementing the Appropriate Access Control Requirements 45 Working with Security over the Network 48 Creating and Managing Azure AI Services 50 Deploying an Azure AI Services Resource 50 Cost Optimization Strategies for Azure AI Services 52 Implementing a Container Deployment 52 Working with APIs and SDKs in Azure 54 Practical: Designing an AI Solution with the REST API 57 Monitoring Azure AI Services 61 Proactively Monitoring Costs 61 Using Metrics and Alerts 62 Practical: Designing Your AI Solution 65 Chapter Review 70 Chapter Quiz 71 3. Storing, Interpreting, and Visualizing Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Data Storage and Management in Azure AI 75 Choosing Storage Options for AI Solutions 76 Data Management Best Practices 84 Data Interpretation for AI Solutions 91 Leveraging Azure AI for Data Analysis 91 Model Training and Selection in Azure AI 93 Data Visualization Techniques and Real-Time Analytics 97 Introduction to Azure Data Visualization Tools 97 Real-Time Analytics and Decision Making 99 Implementing AI in Data Analysis 101 Integrating AI with Azure’s Data Platforms 102 Case Study of Data Analysis in Action 103 iv | Table of Contents

Page 7

Practical: Building an AI-Powered Analytics Dashboard 105 Setting Up Azure Blob Storage 105 Uploading the Customer Feedback Data 106 Creating an Azure AI Services Language Service 106 Creating and Configuring an Azure SQL Database 106 Setting Up Azure Data Factory 107 Creating a Pipeline for Data Movement and Transformation 108 Creating a SQL Database to Store the Data 110 Visualizing the Data with Power BI 111 Chapter Review 112 Chapter Quiz 112 4. Building Decision Support Solutions with Azure AI. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 Introduction to Decision Support for Azure AI 115 Understanding Decision Support Systems 116 What Does Decision Support Look Like in the Azure AI Landscape? 117 Utilizing Azure AI Metrics Advisor to Implement Data Monitoring Solutions 118 Understanding Azure AI Metrics Advisor 118 Implementing Data Monitoring Solutions 120 Text Classification and Moderation with Azure AI Content Safety 127 Understanding Azure AI Content Safety 128 Moderating Text with Azure AI Content Safety 135 Moderating Images with Azure AI Content Safety 139 Detecting Jailbreak Risks 141 Detecting Protected Material 143 Content Filtering for Text Moderation 143 Practical: Implementing a Text Moderation Solution with Azure AI Content Safety 145 Chapter Review 152 Chapter Quiz 153 5. Implementing Computer Vision Solutions with Azure AI. . . . . . . . . . . . . . . . . . . . . . . . 157 Introduction to Azure AI Vision 158 What Is Computer Vision? 158 What Is Azure AI Vision? 159 Case Study 159 Image Analysis with Azure AI Vision 160 The Fundamentals of Image Analysis 161 Performing Image Analysis 162 Extracting Text with Azure AI Vision 169 Facial Recognition and Analysis 172 Table of Contents | v

Page 8

Fundamentals of Facial Recognition 173 Practical: Implementing Facial Recognition in Your Application 176 Custom Vision and Object Detection 178 Building Custom Image Classification Models 178 Implementing Object Detection 182 Practical: Creating a Custom Vision Object Detection Solution 184 Working with Video Content 189 Using Azure AI Video Indexer 190 Practical: Analyzing Video Content with Azure AI Video Indexer 190 Chapter Review 192 Chapter Quiz 193 6. Implementing Natural Language Processing Solutions. . . . . . . . . . . . . . . . . . . . . . . . . . 197 Fundamentals of Natural Language Processing 198 Introduction to NLP 198 Core Components of NLP 198 Common NLP Techniques and Algorithms 199 A Look into NLP in Microsoft Azure 202 Introduction to the Azure AI Language Service 203 Understanding Azure AI Language 204 Using Prebuilt Solutions 205 Using Azure AI Speech to Process Speech 217 Understanding Azure AI Speech 217 Implementing Prebuilt Speech Solutions 218 Implementing Custom Speech Solutions 225 Translating with Azure AI Translator 225 Understanding Azure AI Translator 226 Implementing Prebuilt Translation Solutions 228 Implementing Custom Translation Solutions 232 Practical: Building a Custom Translation Solution 233 More Best Practices 235 Chapter Review 236 Chapter Quiz 237 7. Advanced NLP Techniques and Language Understanding. . . . . . . . . . . . . . . . . . . . . . . . 243 Working with Language Understanding Models 243 Creating Intents, Utterances, and Entities 244 Building Language Understanding Models 246 Optimizing Language Understanding Models 249 Backing Up and Recovering Language Understanding Models 250 vi | Table of Contents

Page 9

Practical: Building and Integrating Your Own Language Understanding Model 251 Building Question-Answering Solutions 259 Understanding Question-Answering Solutions 259 Practical: Building Your Own Question-Answering Solution 262 Advanced Capabilities 267 Chapter Review 278 Chapter Quiz 278 8. Implementing Knowledge Mining and Document Intelligence Solutions. . . . . . . . . . 281 Planning and Implementing a Knowledge-Mining Solution with Azure AI Search 281 Understanding Azure AI Search 282 Creating a Search Service 285 Optimizing Search Performance 288 Maintaining a Search Solution 290 Advanced Search Features 293 Working with Document Intelligence Solutions in Azure AI Document Intelligence 303 Understanding Azure AI Document Intelligence 303 Practical: Implementing a Custom Document Intelligence Model 310 Practical: Creating a Composed Document Intelligence Model 315 Practical: Building Custom Skills for Azure AI Search 317 Chapter Review 323 Chapter Quiz 323 9. Utilizing the Azure OpenAI Service for Generative AI Applications. . . . . . . . . . . . . . . . . 327 Generative AI on Microsoft Azure 327 Types of Generative AI Models 327 Building Responsible Generative AI Solutions 332 Prompt Engineering with the Azure OpenAI Service 336 Writing Effective Prompts 337 Advanced Techniques and Best Practices 338 Generating Content with the Azure OpenAI Service 339 Using the Azure OpenAI Service 339 Using Azure OpenAI in Your Applications 342 Generating Text 343 Generating Images 347 Generating Code 351 Fine-Tuning and Optimizing Generative AI Models 356 Fine-Tuning Your OpenAI Model 357 Table of Contents | vii

Page 10

Integrating Data Sources 358 Interacting with the Model 359 Retrieval-Augmented Generation 360 Understanding RAG 360 Practical: Implementing RAG with Azure OpenAI and Azure AI Search 361 Chapter Review 370 Chapter Quiz 370 10. The Future of AI in Microsoft Azure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 A Look at Key Trends 375 Advances in Complex Reasoning 376 Small Language Models 377 Multimodal AI 380 Agentic AI 382 The Integration of Azure AI with the Entire Azure Platform 382 Practical: Creating a Complex Architecture on Azure with AI Services 384 Streamlining the AI Development Process 396 Developing with Microsoft Copilot 397 Practical: Using Copilot in the Azure Ecosystem (with Code) 398 Working with Prompt Flow in Azure AI Foundry 401 Microsoft Fabric 403 Building AI Applications with Microsoft Fabric 403 Fostering a Data Culture Across Organizations 403 Practical: Creating an End-to-End Solution with Microsoft Fabric 405 A Closing Note 409 Chapter Quiz 409 Appendix: Answer Keys. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413 Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427 viii | Table of Contents

Page 11

Preface Artificial intelligence is rapidly transforming industries, from automated decision making and intelligent processes to deeply personalized customer experiences. Microsoft Azure stands at the forefront of this revolution, offering an end-to-end platform of AI services, machine learning tools, and conversational AI frameworks. But mastering these capabilities and validating that expertise with the certification requires more than memorizing facts. It demands hands-on experience, a clear understanding of real-world challenges, and the confidence to architect solutions that are robust, compliant, and cost-effective. In these pages, you’ll discover a hands-on, narrative approach to learning Azure AI. You’ll begin by exploring how to think like an AI engineer, choosing the right service for a problem, anticipating data challenges, and designing for scale. As you move for‐ ward, you’ll dive into code, creating and customizing models and embedding AI into applications you actually care about. I hope this book will feel like a guide at your side rather than an impersonal manual. When you finish a chapter, you’ll not only have a new skill under your belt; you’ll understand when, why, and how to apply it. This is the journey I’d have wanted to take when I first tackled Azure AI, with clear signposts, real-world context, and the freedom to explore. Whether you’re coding your first bot, operationalizing models in production, or simply curious about what’s possible, I’m confident you’ll find both inspiration and practical know-how here. Why I Wrote This Book Organizations today are investing heavily in AI-powered solutions. As a result, the demand for AI professionals with expertise in designing, deploying, and managing AI solutions on cloud platforms has skyrocketed. Microsoft Azure, with its robust AI services and seamless integration into enterprise environments, has emerged as a leading platform for AI development. ix

Page 12

Yet despite this growing demand, many aspiring Azure AI engineers struggle to navi‐ gate the complexities of the AI-102 certification exam. While Microsoft provides offi‐ cial documentation, there is often a gap between theoretical knowledge and real- world application. That’s why I wrote this book—to bridge that gap and provide a comprehensive, practical guide for passing the AI-102 exam while also equipping readers with skills they can apply beyond the certification. Drawing from my extensive experience in AI engineering, cloud architecture, and exam development, I’ve designed this book to be more than just a study guide. It’s a resource that helps readers master Azure AI Services, implement AI-driven solutions, and prepare for real-world challenges in AI engineering. Whether you’re looking to pass the exam or enhance your AI skills, this book will provide the structured learn‐ ing path you need. Who This Book Is For This book is intended for a wide range of readers who are looking to become profi‐ cient in Azure AI and earn the AI-102 Azure AI Engineer Associate certification. It will be particularly useful for: • Aspiring AI engineers who want to build a solid foundation in designing and deploying AI solutions on Azure • Software developers and cloud engineers looking to expand their expertise in AI and machine learning • Data scientists and analysts who want to leverage Azure AI services to develop intelligent applications • IT professionals and solution architects who need to understand AI integration and security best practices within the Azure ecosystem • Certification candidates who want a structured approach to passing the AI-102 exam with confidence This book assumes a basic understanding of cloud computing and Python program‐ ming. However, no prior AI expertise is required; concepts are introduced progres‐ sively, making it accessible for beginners while still being valuable for experienced professionals. How This Book Is Organized This book follows a structured approach, covering all six domains of the AI-102 cer‐ tification exam while providing practical, hands-on experience with Azure AI solutions. x | Preface

Page 13

Chapter 1, “Introduction to AI Solutions on Microsoft Azure” provides an overview of the AI landscape, responsible AI considerations, and technical setup for Azure AI development. Chapter 2, “Planning and Managing AI Solutions in Microsoft Azure” covers the AI project lifecycle, security and access management, deployment strategies, and moni‐ toring AI services. Chapter 3, “Storing, Interpreting, and Visualizing Data” explores data storage, analy‐ sis, and visualization techniques to support AI-driven decision making. Chapter 4, “Building Decision Support Solutions with Azure AI” details how to use Azure Cognitive Services for decision making, anomaly detection, and personalized recommendations. Chapter 5, “Implementing Computer Vision Solutions with Azure AI” walks through image analysis, facial recognition, object detection, and video content processing. Chapter 6, “Implementing Natural Language Processing Solutions” introduces NLP fundamentals, text classification, and Azure AI Language services. Chapter 7, “Advanced NLP Techniques and Language Understanding” covers custom NLP models, named entity recognition, and conversational AI applications. Chapter 8, “Implementing Knowledge Mining and Document Intelligence Solutions” focuses on Azure AI Search, document intelligence, and search optimization. Chapter 9, “Utilizing the Azure OpenAI Service for Generative AI Applications” explains generative AI workflows, prompt engineering, fine-tuning models, and retrieval-augmented generation (RAG). Chapter 10, “The Future of AI in Microsoft Azure” discusses emerging AI trends, Microsoft Fabric, and streamlining AI development with Copilot. Each chapter includes hands-on exercises, quizzes, real-world scenarios, and exam- focused insights to ensure readers gain both practical and theoretical expertise. The answers to chapter quiz questions can be found in the Appendix. Conventions Used in This Book The following typographical conventions are used in this book: Italic Indicates new terms, URLs, email addresses, filenames, and file extensions. Preface | xi

Page 14

Constant width Used for program listings, as well as within paragraphs to refer to program ele‐ ments such as variable or function names, databases, data types, environment variables, statements, and keywords. Constant width bold Shows commands or other text that should be typed literally by the user. Constant width italic Shows text that should be replaced with user-supplied values or by values deter‐ mined by context. This element signifies a tip or suggestion. This element signifies a general note. This element indicates a warning or caution. Using Code Examples Supplemental material (code examples, exercises, etc.) is available for download at https://oreil.ly/azure-ai-engineer-associate-study-guide-supp. If you have a technical question or a problem using the code examples, please send email to support@oreilly.com. This book is here to help you get your job done. In general, if example code is offered with this book, you may use it in your programs and documentation. You do not need to contact us for permission unless you’re reproducing a significant portion of the code. For example, writing a program that uses several chunks of code from this book does not require permission. Selling or distributing examples from O’Reilly books does require permission. Answering a question by citing this book and quoting example code does not require permission. Incorporating a significant amount of xii | Preface

Page 15

example code from this book into your product’s documentation does require permission. We appreciate, but generally do not require, attribution. An attribution usually includes the title, author, publisher, and ISBN. For example: Azure AI Engineer Associate (AI-102) Study Guide by Renaldi Gondosubroto (O’Reilly). Copyright 2025 Renaldi Gondosubroto, 978-1-098-16926-8.” If you feel your use of code examples falls outside fair use or the permission given above, feel free to contact us at permissions@oreilly.com. O’Reilly Online Learning For more than 40 years, O’Reilly Media has provided technol‐ ogy and business training, knowledge, and insight to help companies succeed. Our unique network of experts and innovators share their knowledge and expertise through books, articles, and our online learning platform. O’Reilly’s online learning platform gives you on-demand access to live training courses, in-depth learning paths, interactive coding environments, and a vast collection of text and video from O’Reilly and 200+ other publishers. For more information, visit https://oreilly.com. How to Contact Us Please address comments and questions concerning this book to the publisher: O’Reilly Media, Inc. 141 Stony Circle, Suite 195 Santa Rosa, CA 95401 800-889-8969 (in the United States or Canada) 707-827-7019 (international or local) 707-829-0104 (fax) support@oreilly.com https://oreilly.com/about/contact.html We have a web page for this book, where we list errata, examples, and any additional information. You can access this page at https://oreil.ly/azure-AI-engineer-associate- AI-102-study-guide-1e. For news and information about our books and courses, visit https://oreilly.com. Find us on LinkedIn: https://linkedin.com/company/oreilly-media Watch us on YouTube: https://youtube.com/oreillymedia Preface | xiii

Page 16

Acknowledgments Writing this book was a journey made possible by the support and inspiration of many individuals. First and foremost, thank you to the O’Reilly team for being not just professional partners, but a genuine pleasure to work with, and for entrusting me with this project. A special shout-out to Angela Rufino, my content development edi‐ tor: your insight and guidance were invaluable at every turn. I’d also like to acknowl‐ edge Ashley Stussy, Doug McNair, Kristen Brown, Megan Laddusaw, Rachel Wheeler, and Judith McConville whose thoughtful support and expertise consistently helped throughout the development process. I couldn’t have asked for a more dedicated or talented group to bring this book to life. I extend my gratitude to my colleagues, mentors, and peers in the AI and cloud com‐ puting industry who provided valuable feedback and discussions that enhanced the depth of this book. Your expertise and encouragement were instrumental in refining the material. My sincere thanks go to the technical reviewers of this book—Prashanth Chaitanya, Vaibhav Gujral, Shashank Pawar, and Rebeca Whitcomb—whose careful feedback has ensured the highest quality of this book. I’m also grateful to the AI and cloud com‐ munities, especially those in the Microsoft Azure ecosystem, for their ongoing spirit of innovation and collaboration. The discussions and challenges shared in user groups, conferences, and online forums have been instrumental in shaping the con‐ tent you’re about to read. A special thank you goes to my family and friends, whose unwavering support and patience allowed me to dedicate the countless hours needed to complete this project. Their encouragement kept me motivated throughout the writing process. Finally, I’d like to acknowledge the readers of this book; your dedication to learning and professional growth is what drives me to create resources like this. I hope this guide empowers you to succeed in your AI-102 certification journey and beyond. xiv | Preface

Page 17

CHAPTER 1 Introduction to AI Solutions on Microsoft Azure The AI gold rush is here—but instead of pickaxes, everyone’s scrambling for GPUs and prompt engineering skills. Last quarter, a retail client told me their board deman‐ ded “generative AI something” by Friday…but their team was still manually tagging product images. That’s the chaos driving the AI-102 certification’s surge: companies aren’t just chasing ChatGPT headlines—they’re desperate for engineers who can turn Azure’s toolbox into actual business wins. Therefore, there’s no better time than now to take the AI-102 exam and get certified for your skills in working with AI solutions on Microsoft Azure. It’ll help boost your credibility as an AI engineer and set you apart in this job market, where there’s huge demand for personnel who possess AI skills. In this chapter, we’ll examine the current AI landscape and explore what will be expected from you as an AI-102 exam candidate. I’ll also help you prepare your Microsoft Azure environment for usage in developing solutions and gain an under‐ standing of Azure AI’s capabilities. I recommend that you complete the AI-900 certification before beginning this study guide (though it’s not mandatory for you to do so). This guide aims to walk you through getting certified from start to finish, and having exposure to common Micro‐ soft Azure AI concepts will be helpful as you work through it. 1

Page 18

1 John McCarthy, “The Dartmouth Workshop—As Planned and as It Happened”, October 30, 2006. Introduction to the AI Landscape Picture this: it’s 1956, and a handful of scientists at a Dartmouth College workshop1 are huddled around punch cards, dreaming of machines that “think.” Back then, building AI was like writing out every step of a recipe for someone who’s never cooked before, from how to crack an egg to when to stir, leaving no room for improv‐ isation. Fast-forward to the 1990s, and machine learning flipped the script. Suddenly, instead of handcoding how to spot a cat in a photo, we let algorithms binge-watch thousands of images until they figured it out themselves. And now, in the 2020s, working with Azure AI feels less like babysitting those early rulebooks and more like collaborating with a partner who’s read every manual ever written and somehow stayed awake through it all. What was Microsoft’s big play? Consolidating its AI tools under one roof with Azure AI Foundry. Think of it as your AI workshop. Need a quick text translator? Grab the prebuilt Azure AI Translator off the shelf. Want to craft a custom chatbot that sounds like your CEO? Fire up GPT-4 in Azure OpenAI in the same workspace. But here’s the catch I’ve seen trip up teams: choosing between these tools isn’t about “advanced versus basic”―it’s like choosing between a power drill and a Swiss Army knife. Let’s say you’re building a medical app. A fine-tuned Azure AI Vision API could ana‐ lyze X-rays out of the box, while Azure OpenAI’s models could generate patient sum‐ maries that even your time-crunched nurses would trust. What the docs won’t tell you is that fine-tuned vision API might cost three times more per scan than a production-ready model—a trade-off that keeps CFOs up at night. Yet when rare conditions require detection of subtle image markers, or when regulatory compliance demands explainable, domain-specific insights, it can be worth the extra expense to fine-tune a model so patients receive more accurate diagnoses and hospitals avoid costly errors. Additionally, while representational state transfer (REST) APIs remain the primary interface for many Azure AI offerings, developers may benefit from an understanding of asynchronous (async) programming patterns, such as those they may use in async/ await in Python when trying to handle long-running AI operations successfully. Familiarity with Azure fundamentals such as resource groups, networking, and cost management will also be increasingly critical for developers as AI solutions scale into enterprise settings. In this section, I’ll introduce what is expected from you as a candidate taking the AI-102 exam, where we are now with AI, and the fundamental concepts you’ll need 2 | Chapter 1: Introduction to AI Solutions on Microsoft Azure

Page 19

to know to understand AI’s usage within the Microsoft Azure environment for the AI-102 certification. The Minimally Qualified Candidate for the AI-102 Exam Think of the AI-102 as your “commercial driver’s license” for Azure AI—it’s where you prove you can haul real-world AI solutions, not just cruise around with theory. It builds on the AI-900, which is the foundational certification test that focuses on fun‐ damental machine learning and AI concepts that are within the Microsoft Azure eco‐ system. Candidates are expected to be proficient in the different phases of AI solution development, which include requirements definition, design, development, deploy‐ ment, integration, maintenance, and monitoring. This makes the AI-102 a highly interdisciplinary exam that requires you not only to understand the development process of AI systems but also to be familiar with other aspects of the system lifecycle. To prepare yourself to take the exam, you will need to gain an understanding of the services that make up the Azure AI portfolio and know when to apply which service to which situation. This will include gaining an understanding of the data sources that you will be working with for the AI services you will utilize. You must also abide by the principles of responsible AI that Microsoft has established, be able to use REST APIs and software development kits (SDKs) to consume the Azure AI services, and have the REST APIs and SDKs integrated with applications in your own environ‐ ments. We will examine this further later in the chapter. Candidates also need to be comfortable with at least one language that’s used to work with AI solutions. Throughout this book, we will be using Python for developing our AI solutions. It’s a popular choice among AI engineers due to its extensive library ecosystem―it has SDKs available for many different AI services from major cloud providers and third parties. Additionally, its simplicity and readability make it an easy language to learn, write, and understand, which are essential characteristics for any language that’s used to write the complex algorithms used in AI. While Python is widely used, other languages, such as R (which excels at statistical modeling and data visualization) and Java (which offers robust libraries for production-scale AI sys‐ tems), also play important roles in AI development. If you don’t have much exposure to Python, I recommend that you consult some beginner-level documentation at the official Python website, take some beginner-level tutorials there, and learn a bit more about the use cases of the language to help you understand the foundational syntax we will be using throughout this guide. This guide is meant not only to help you get through the exam, but also to provide you with valuable skills that you can apply to your work as an AI engineer. It will help Introduction to the AI Landscape | 3

Page 20

2 International Data Corporation, “IDC: Artificial Intelligence Will Contribute $19.9 Trillion to the Global Economy Through 2030 and Drive 3.5% of Global GDP in 2030”, September 17, 2024. you demonstrate your skillset practically in the industry, where there’s currently a high demand for AI skills. Candidates for the AI-102 exam should also be prepared to understand and configure role-based access control (RBAC) within Azure. For instance, Azure provides built-in roles such as “Cognitive Services Contributor” and “Cognitive Services Reader.” You may also create custom roles to manage access more precisely. For example, an enter‐ prise might allow only certain data scientists to deploy models, while giving a broader group of analysts permission to run or test those models. Understanding how these permissions work in typical enterprise scenarios, such as controlled development environments and production deployments, will help you ensure that your AI solu‐ tions remain both secure and compliant. Now that you have an understanding of what you need for the exam, we can move on to taking a look at the current state of AI. Where We Are Now with AI Today, AI is a rapidly growing field that has applications in almost every industry. From healthcare to finance, we can see AI being used in automating tasks, making predictions, and improving day-to-day decision making. To an extent, it has become advanced enough that it has surpassed human performance in certain areas, such as image recognition and natural language processing (NLP). Given expectations that AI will contribute $15.7 trillion to the global economy by 2030,2 it’s no surprise that the demand for AI professionals is continuously rising, with organizations looking for skilled individuals to implement AI solutions. Various providers offer these AI technologies, whether in the form of specific solutions or centralized platforms hosted in the cloud. Using platforms within the cloud has made it easier and more affordable for many to leverage the advantages of AI. Previously, training your own model could be very expensive for the average organization. The rise of cloud-based AI platforms such as Amazon Web Services, Microsoft Azure, and Google Cloud has eliminated the need for users to invest in expensive hardware and infrastructure. One of the leading fields in AI is generative AI, which is artificial intelligence that can create new and original content, including images, video, music, and text. It has many potential applications, such as creating realistic marketing campaign plans, virtual environments for video games, and new forms of art and music. We will discuss how to leverage the full potential of generative AI in Microsoft Azure in Chapter 8 of this guide. 4 | Chapter 1: Introduction to AI Solutions on Microsoft Azure