Managing AI Projects Drive Innovation and Successfully Navigate the Full AI Project Lifecycle (Malini Jain Runtasewee etc.)（Z-Library）

Managing AI Projects Drive Innovation and Successfully Navigate the Full AI Project Lifecycle MALINI JAIN RUNTASEWEE & ADRIÁN GONZÁLEZ SÁNCHEZ

Successfully delivering AI projects requires more than technical expertise—it demands a new kind of project management (PM). Managing AI Projects is your practical guide to leading AI initiatives from idea to production. Written by seasoned experts Malini Jain Runtasewee and Adrián González Sánchez, this book blends traditional PM principles with the realities of AI development, helping you structure projects, manage uncertainty, and drive real outcomes. Whether you’re a project manager, engineer, or product lead, you’ll learn how to plan AI initiatives, support iterative experimentation, and align technical and business teams. With practical tools, real-world examples, and a focus on both traditional and generative AI, this book helps ensure your AI projects deliver impact beyond just initial pilots and prototypes. • Structure AI projects from ideation through deployment • Manage uncertainty, experimentation, and changing requirements • Bridge technical and nontechnical teams effectively • Reduce risk and increase success using proven practices • Deliver AI initiatives that align with business goals and timelines • Develop an applied AI project management handbook for your day-to-day initiatives Malini Jain Runtasewee is a senior cloud and data project manager and a university lecturer with IE University and EOI. She is also coauthor of an AI fundamentals book published by ANAYA Multimedia. Adrián González Sánchez is an AI product manager at Microsoft AI and an author of multiple O’Reilly books and reports. He trains professionals at HEC Montreal and IE University, and has authored online courses for LinkedIn Learning, O’Reilly, The Linux Foundation, and DeepLearning. AI. PROJEC T MANAGEMENT “In an era of rapid technological evolution, this book is your roadmap to keeping strategy, people, and technology moving in sync. It’s become my new reference guide for implementing AI projects.” — Jason Nitz, mining technical and operations consultant “A must-read for any project manager responsible for turning AI ambition into real world outcomes.” — Sheldon Rodrigues, senior applied scientist, Microsoft Managing AI Projects DRIVE INNOVATION AND SUCCESSFULLY NAVIGATE THE FULL AI PROJECT LIFECYCLE ISBN: 979-8-341-64101-3 US $44.99 CAN $56.99

Praise for Managing AI Projects Packed with practical insights, proven frameworks, and real‑world examples, Managing AI Projects demystifies the complexity of delivering a successful AI implementation. It shows you how to align teams, manage risk, and translate technical innovation into measurable impact. In an era of rapid technological evolution, Managing AI Projects is your roadmap to keeping strategy, people, and technology moving in sync. It’s become my new reference guide for implementing AI projects. —Jason Nitz, mining technical and operations consultant This book is an excellent resource treating AI project management as a serious discipline. It combines technical realism, delivery pragmatism, and organizational insight in a way that most other AI books avoid. A must-read for any project manager responsible for turning AI ambition into real world outcomes. —Sheldon Rodrigues, senior applied scientist, Microsoft

Managing AI Projects Drive Innovation and Successfully Navigate the Full AI Project Lifecycle Malini Jain Runtasewee and Adrián González Sánchez

979-8-341-64101-3 [LSI] Managing AI Projects by Malini Jain Runtasewee and Adrián González Sánchez Copyright © 2026 Adrián González Sánchez and Malini Jain. 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/institutional sales department: 800-998-9938 or corporate@oreilly.com. Acquisitions Editor: David Michelson Development Editor: Angela Rufino Production Editor: Christopher Faucher Copyeditor: Paula L. Fleming Proofreader: J.M. Olejarz Indexer: BIM Creatives, LLC Cover Designer: Susan Thompson Cover Illustrator: Susan Thompson Interior Designer: Monica Kamsvaag Interior Illustrator: Kate Dullea May 2026: First Edition Revision History for the First Edition 2026-05-19: First Release See https://oreilly.com/catalog/errata.csp?isbn=9798341641013 for release details. The O’Reilly logo is a registered trademark of O’Reilly Media, Inc. Managing AI Projects, the cover image, and related trade dress are trademarks of O’Reilly Media, Inc. The views expressed in this work are those of the authors and do not represent the publisher’s views. While the publisher and the authors have used good faith efforts to ensure that the infor- mation and instructions contained in this work are accurate, the publisher and the authors dis- claim 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 instruc- tions 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.

Contents | Preface: Our AI Project Management Dream vii 1 | Introduction to AI Project Management 1 2 | A Deep Dive into AI for Project Managers 31 3 | The Role of the AI Project Manager 73 4 | Applied Approach to AI Project Management 117 5 | PM Considerations During the Technical AI Lifecycle 171 6 | Tools for Managing AI Projects 221 7 | Tales of AI Project Innovation 251 | Index 269 v

Preface: Our AI Project Management Dream Here we are. The fourth O’Reilly book after my first three titles: Kubernetes and Cloud Native Associate (KCNA) Study Guide, Azure OpenAI Service for Cloud Native Applications, and Generative AI on Microsoft Azure. All of them focus heav- ily on cloud computing and artificial intelligence and are fairly technical. As an author, I like to think these books help people who are not necessarily hypertech- nical understand complex topics and join the tech industry. Due to personal bias and writing style, I believe some of these books are great resources for software architects and technical product managers, who are part of that privileged cohort of hybrid professionals with mastery of both technical and business topics, espe- cially in this era of artificial intelligence. Also part of that group are traditional project managers (PMs)—or they can be, at least. Often misunderstood and confused with other roles that accompany Agile methodologies, like scrum masters and product owners, PMs are among the tactical professionals who manage and facilitate work at different levels, including projects, programs, and portfolios. Yet regardless of their role in the organization or their job title, PMs have taken longer than many others to join the wave of AI learning and upskilling. Don’t get me wrong: there are wonderful PMs working on data and AI projects who possess a relatively high level of technical knowledge and the ability to help all sorts of technical professionals perform several times better. But it is also true that we see, for example, more product managers than project managers with appropriate levels of AI understanding, likely due to the fact that AI adoption has arisen mostly in product-led organizations. This is changing, though, with more kinds of organizations adopting AI and including it vii

in their projects. The trend directly impacts the project management professio- nals who are leading and facilitating AI projects. Nonetheless, the increasing number of AI projects hasn’t yet resulted in a larger number of people who want to learn and apply AI project management. There are plenty of resources dedicated to “AI for project managers,” when this is defined as a set of AI tools to help PMs increase their productivity and manage their projects more efficiently. However, there is very limited research and few resources (including books and courses) that focus on the details of how these professionals can manage and lead their projects to a successful end. As part of my personal mission to alter this landscape, I started teaching AI project management topics in 2019 at multiple institutions in Canada (Concor- dia University and HEC Montréal) and Spain (IE University and the ethical AI observatory OdiseIA). I then expanded on the course content and decided to bring it to the O’Reilly online training platform, with recurrent sessions intended to demystify the specifics of managing AI projects. These sessions included live demonstrations of AI project roadmaps with definitions of tasks, estimates, and assignments to different team members. I regularly collected feedback from course attendees and started to observe a pattern of project management profes- sionals having no or elementary knowledge of AI but a great interest in incorpo- rating AI knowledge into their new AI projects. And, of course, I paid special attention to neutral or negative feedback to understand my students’ key concerns and needs. From those comments, I understood that once the AI project lifecycle terms were clear, a lot of project management topics related to Agile methodologies and operational concerns, such as planning, estimating, and delivering projects on time, were similar to those of any other project they were managing. I got very positive feedback from the live demonstrations, because the PMs could get solutions for challenges they had never faced before. I confirmed this point during other professional coaching and tutoring activities with the participants in the MIT Sloan/CSAIL GetSmarter course, in which we had to build and evaluate AI project plans for multiple industries and thousands of organizations. My coauthor for this book, Malini Jain Runtasewee, complements this back- ground by bringing a very rare skill set that includes not only the ability to man- age complex projects for cloud, data, and AI implementations but also the capacity to get the best performance from varied different team members. I say “rare” because she is able to continue evolving her playbook project after project without being perceived by teammates as manipulative or bossy. She guides, viii | PREFACE

coaches, and helps everyone understand their role and key expectations while connecting points, extracting dependencies, and anticipating risks that are highly specific to the data and AI roles and activities of the project. That said, the challenge for this book was clear: how to strike a balance between the AI and project management fundamentals and the techniques and terms that an AI project manager needs to know. This book addresses exactly that. These chapters cover a mix of topics that will enable you, our dear reader, to leverage best practices with real and immediate value, as well as tools and tem- plates for your project-planning activities. No matter what your exact project management background is, you will find this book not only easy to understand but also specific enough to develop your body of knowledge in AI project management. In this book, you will learn our EMED (exploration, mobilization, execution, and delivery) methodology for AI project management. This approach, based on our combined experience and then tested and improved over the years, applies to both our extensive experience with managing technology projects and our experi- ence with AI projects in companies around the world. And this book is only the core piece of a wider learning experience, which also includes new O’Reilly learn- ing paths and live online trainings, as well as other courses on LinkedIn Learn- ing, DeepLearning.ai, and other platforms. We hope you find this book informative. We also hope you enjoy it. This book is a very personal project, one that draws on many professionals’ stories, ongoing discussions with friends and colleagues, and a vision of how project management best practices and practitioners need to evolve, upskill, and increase their level of specialization to bring tangible value to AI projects and teams. It is an obvious statement nowadays, but it took time for the project management field to understand and adopt AI and to embrace this new era of advanced, tacti- cal AI management. —Adrián González Sánchez PREFACE | ix

Who This Book Is For This book is a practical guide to AI project management. Instead of simply com- bining AI and traditional project management concepts, it goes deep into project and technical lifecycles and the techniques required to successfully implement AI initiatives from initial ideation through final delivery. For that reason, this book is a great resource for: • Experienced project managers who need practical frameworks to plan, scope, de-risk, and deliver AI projects in real-world environments • Scrum masters and product owners who want to adapt Agile practices to the uncertainty, experimentation, and data dependency of AI work • Technical professionals (e.g., data scientists, machine learning [ML] engi- neers, AI engineers, developers) who want to better understand gover- nance, stakeholder management, delivery structures, and how their work fits into larger business outcomes • Other tactical managers (e.g., product, operations, and IT managers; innova- tion and transformation leads) responsible for executing AI initiatives and coordinating cross-functional teams • Executives and managers looking for guidance on hiring and needing clarity on processes, roles, team design, lifecycle governance, etc. • Coaches and consultants who support organizations through AI transforma- tions and need structured, field-tested delivery models to guide clients effectively • Entry-level professionals and career switchers who seek a practical, structured introduction to how organizations’ AI projects are actually run x | PREFACE

How This Book Is Organized This book presents an end-to-end approach to AI project management. Organ- ized into seven chapters (see Figure P-1), it offers all the knowledge you will need to succeed in your job and, if desired, to land another position in today’s competi- tive job market. It also provides plenty of notes, advice, and answers to questions like “Why is this relevant for AI project managers?” and “How is this different from regular project management?” Moreover, you’ll appreciate the incremental approach we take, allowing you to build your knowledge step-by-step. Figure P-1. This book’s topics by chapter Chapter 1, “Introduction to AI Project Management” Offering far more than the usual primer, this chapter will go into the details of AI project management, including the reasons why it has become a new domain of work, the key considerations involved in its practice (espe- cially compared to regular project management), and the main challenges and concerns PMs need to be aware of when dealing with AI projects. This chapter sets the stage for the rest of the book, establishing foundational knowledge of the topic, and will be a useful basis for moving forward regardless of your AI, project management, and AI project management experience. Chapter 2, “A Deep Dive into AI for Project Managers” Because we believe that no PM can manage AI projects without certain baseline AI knowledge, this chapter gets more technical, explaining key terms and technologies. It then connects this information to AI project management topics such as anticipating potential risks related to new and less mature technologies, the impact of technology stack choice on the pace PREFACE | xi

of implementation and total project duration, and how the feasibility of specific use cases depends on the type of AI technology used. In summary, Chapter 2 will contribute to your evolution toward being a “technical project manager.” Chapter 3, “The Role of the AI Project Manager” The third chapter covers the role of the AI project manager and the rela- tionship of the AI PM with every other stakeholder and role, including technical team members. It explores all the considerations involved in managing AI teams and discusses techniques specific to the context of data and AI teams, taking a PM-centric perspective on the topic. Chapter 4, “Applied Approach to AI Project Management” This chapter discusses our distinctive approach to AI project management, from the atypical AI project management lifecycle stages that we use in our day-to-day work to ways of combining project management methodologies you may be familiar with. With a focus on risk anticipation and mitigation, this chapter will prepare you for the technical aspects of AI lifecycles in the next chapter. Chapter 5, “PM Considerations During the Technical AI Lifecycle” This chapter brings the perspective of AI project management to each stage of the project implementation lifecycle. In fact, this is the second technical deep dive in this book, taking you through a series of stages and more- granular steps, while keeping in mind the role of the AI project manager and its importance to the progress of the entire team. Chapter 6, “Tools for Managing AI Projects” This chapter explains the technical toolkit for both managing and imple- menting AI projects. We discuss how to use the PM tools you may already be familiar with for AI projects, and we’ll cover other technical tools you need to know how to use for your work with technical teams. Chapter 7, “Tales of AI Project Innovation” The last chapter contains a series of project experiences, illustrative exam- ples, and applied recommendations that will help bring to life the situa- tions you may face in the complex reality of AI projects. xii | PREFACE

In summary, the topics covered in these chapters address everything you need to upskill and shine as an AI project manager, positioning you to differenti- ate your skill set and the value you can add to data and AI teams. O’Reilly Online Learning For more than 40 years, O’Reilly Media has provided tech- nology 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 and any additional information. You can access this page at https://oreil.ly/managing-AI-projects. For news and information about our books and courses, visit https:// oreilly.com. Find us on LinkedIn: https://linkedin.com/company/oreilly. Watch us on YouTube: https://youtube.com/oreillymedia. PREFACE | xiii

Acknowledgments This book represents an extraordinary personal and collective effort to put together years of experience, customized frameworks and models, and a bunch of best practices and learnings from all kinds of mistakes and challenges in the field. At a personal level, it is a very special project because we got the opportunity to work together. Married in real life, we are partners in crime for this writing adventure. So we have to thank each other for the passion, knowledge, and effort we each dedicated to make this book a reality. Even more important, we are very thankful for the continuous support and love of our family. Thank you all for everything—we are very lucky. We must give a shout-out to the amazing work from the entire O’Reilly team. Thanks, David, for believing in this project. Thank you, Angela and Melissa, for your work during the entire writing process. Thanks to the produc- tion, design, and illustrations teams—you folks know how to convert raw text into an amazing-looking book. Thank you all, technical reviewers, for putting real effort into reading and analyzing every single detail of the final manuscript. And finally, thank you, everyone, for all the kind notes, private messages, questions, and pieces of advice. Every time one person tells us that the book is making their AI projects a bit better and more successful, we feel the effort of creating this book was all worth it. This is for you, Khun Mae Sukon Runtasewee. Thank you for everything. We miss you. xiv | PREFACE

Introduction to AI Project Management A few years ago, AI project management wasn’t really a thing. There are several flavors of projects involving AI, machine learning, and related technologies, but managing AI projects didn’t seem to be an area of study for project managers. Somehow, no one thought it would require new techniques or knowledge...until companies started to adopt AI and experience project failure. You probably heard the stories: unjustified return on investment (ROI), only 20% of projects suc- ceeding (based on statistics from Gartner and others), eternal cycles of AI proof- of-conceptionitis (or being afraid to actually move to production and stop the experimentation phase). Implementing AI is not easy. It requires specialization at all levels, from technical roles to executive stakeholders, including all tactical roles at the project, product, program, and portfolio levels. But still, you may be asking: Why focus on the project manager (PM)? Why is this role so important for the art of managing AI projects? This chapter will explain how PMs contribute to every aspect of AI projects. The following list pre- views how the role interacts with other project stakeholders: Executive sponsors Unlike other tactical managers, project-level professionals focus on the pri- oritization, planning, implementation, and completion of AI projects. That means that PMs are the point people for sharing relevant news with any executive who wants to understand the progress of their AI projects. As a team member who sits between the AI delivery team and the executive sponsors, an AI project manager can serve as a translator, or interface, between business and technical stakeholders. 1 | 1

Technical teams Project managers and people in related Agile roles, such as scrum masters and product owners, spend a fair amount of time with the technical teams. There is a relational aspect to PM work that relies on having some empathy and emotional intelligence. The PM reads the day-to-day dynamics, coaches team members, understands potential blockers, finds a way to get addi- tional resources, and adapts their leadership approach to the science, engi- neering, development, or design background of the technical team members. Clients and partners Due to the hybrid and multidisciplinary nature of their role, AI project managers are uniquely positioned to share project progress with relevant stakeholders, calibrating their reports to include an appropriate level of detail. This means sharing concrete information about project progress, for example, during and after project sprints. Key to success as a communica- tor is developing your hybrid profile by increasing your level of technical knowledge while optimizing your AI PM approach. (Good news—that is the goal of this book!) Other tactical managers While AI PMs focus on project details and progress, other professionals— such as product, program, and portfolio managers—work at different levels in the organization. They may handle AI products or a set of projects within a program/portfolio, but their type and scope of action is different. You, as an AI project manager, can bring concrete project insights, antici- pate day-to-day roadblocks, spot needs for new resources, and so forth. While these tasks are not specific to AI (for example, check out this report from the Project Management Institute on collaboration between project and product professionals), this level of collaboration will be very relevant if your organization has these roles and you need to coordinate and collabo- rate with them. These collaborative workstreams are closely tied to how we manage AI within organizations. Managing them requires a multilevel approach that com- bines top-down and bottom-up initiatives. Let’s now explore the various levels of AI management. 2 | MANAGING AI PROJECTS

Levels of AI Management All the stakeholders that you’ll interact with as an AI project manager are related to the trifecta of strategic, tactical, and technical/operational AI management within the company, as you can see in Figure 1-1. Figure 1-1. Levels of AI management Let’s discuss each level in more detail as a way to examine how your role con- tributes to the different levels of your company’s AI management. LEVEL 1: AI STRATEGY The top level is top-down and strategic, aligning AI with the organization’s over- all business strategy. The AI strategy should lay the foundation for properly implementing AI technologies, based on a series of strategic pillars. Each of these AI maturity pillars establishes a top-down framework for embedding AI across the enterprise. Business strategy The business strategy defines how AI can align with organizational goals and cre- ate sustainable value. It emphasizes a long-term vision and objectives for AI adoption. Key objectives include improving operational efficiency, enhancing customer satisfaction, unlocking new revenue streams, and gaining a competi- tive edge. INTRODUCTION TO AI PROJECT MANAGEMENT | 3

Technology and data strategy This area focuses on the infrastructure and data management strategies needed to enable effective AI solutions. It involves selecting the right technologies, man- aging data pipelines, and integrating systems. For example, it may be important to choose and deploy cloud platforms that are capable of handling large data vol- umes and training complex AI models; to build robust pipelines for data process- ing and analysis to ensure accessibility, cleanliness, and quality; or to leverage big data technologies and distributed storage to integrate AI into broader enterprise systems. AI expertise and development strategy This pillar involves the development and practical implementation of AI models within the organization. It emphasizes continuous experimentation, alignment with business objectives, technical feasibility, ROI, and impact on business pro- cesses. There are two main workstreams that contribute to the continuous increase of internal AI expertise. Organizational culture It is important to consider the cultural and organizational readiness to adopt AI, including the workforce’s degree of technical and change management skills and openness to innovation and experimentation. Some potential actions include launching internal training programs to upskill employees in AI; fostering inno- vation and encouraging openness to technological change through hackathons, innovation labs, and incentives for new ideas; and leveraging multidisciplinary teams or pods in which AI experts work side by side with domain specialists to foster collaboration and knowledge sharing. Internal AI governance A good governance program ensures that AI is used ethically and responsibly within the organization. Some of the key elements of proper end-to-end AI gover- nance are these: Ethical AI principles A good first step is to establish some AI principles that are aligned with the organization’s values. These principles function as a declaration of intentions. The Principled Artificial Intelligence initiative, developed by the Berkman Klein Center for Internet & Society at Harvard University, includes a comprehensive collection of references to existing AI principles 4 | MANAGING AI PROJECTS