Introduction

For the past 30 years, I have had the privilege of working in the data and analytics space. I have been at the center of the field's evolution-from relational databases and data mining to the transformative power of artificial intelligence (AI) that we see today.

I began my data career at AT&T Bell Laboratories, working in a Center of Excellence that brought together experts from diverse fields-economists, computer scientists, operations researchers, statisticians, geophysicists (like me), and even psychologists. Our mission was clear: to find innovative ways to solve business problems using data. One example was pioneering the continuous audit of online systems, a concept that is now foundational in modern auditing and compliance. I also had the opportunity to work with Bell Labs research teams to bring early machine learning models into practical business use. At the time, we could predict customer churn with decent accuracy, but deploying those models in production was a different story. Operational constraints made it nearly impossible.

Since then, I have spent more than two decades as a practitioner, industry analyst, thought leader, and researcher. I've had the opportunity to help organizations navigate and capitalize on the most transformative waves, from the rise of e-commerce and the dot-com bubble to the cloud computing revolution, big data, and now AI. I've been an industry analyst. I've coauthored books on cloud computing and big data. For the past 12 years I've worked with a wonderful team at TDWI, a data and AI education and research firm, as the vice president of research.

A Generational Leap in AI

We are at a unique moment in technological history. The advent of generative AI represents a generational leap. It is a transformation as important as the emergence of the internet or mobile computing. Until recently, organizations followed a relatively predictable analytics maturity curve: they started with reports, moved to dashboards, then self-service analytics, and eventually developed predictive models. As organizations became more advanced, they put machine learning models into production for use cases such as churn prediction, fraud detection, recommendation engines, and predictive maintenance.

Not surprisingly, as organizations see tangible value from AI, demand continues to grow. I've observed a clear pattern: the more mature an organization is in its analytics capabilities, the more measurable value it derives from AI. This creates a virtuous circle; as companies benefit from AI, they become more committed to advancing their capabilities further.

But now, generative and agentic AI are trying to change the game. Organizations that were once methodically advancing through the analytics maturity model are now feeling intense pressure to implement AI-driven solutions immediately. Yet, many companies are not yet ready to utilize some of these advances to the fullest. While they think it may be easy to deploy a chatbot for instance, they don't realize that their data foundation needs to be in place if they hope to make use of customer data. They haven't necessarily thought about what resources would be needed to build AI applications at scale. They haven't thought about how to govern their data, their analytics, or their AI. What I see in my research is that many companies are still working to put the data foundation in place (although they are making progress), democratize business intelligence (BI) to make it easier for business users to access and analyze data, and try to govern what they have. And now they are being asked to implement AI!

Much of this hype, of course, is being driven by generative AI. And there are at least two sides to the generative AI coin. There is the consumerization of AI because of generative AI. That means people in companies making use of generative AI platforms (such as ChatGPT, Perplexity, Claude, or Microsoft Copilot) to make their work more efficient. Then there is the flip side, the side I think about as a data and analytics professional; that is building applications that use this technology, against company data, for competitive advantage. Of course, in between is the movement to infuse AI into applications across the data and analytics life cycle. It is a complex time.

The Purpose of This Book

There have been many books written about succeeding with AI-what companies are doing, what their leaders say work, but they are at the leadership level. This book goes down a level, into the technology, into governance, into the organizational structure. It will give you a window into what goes on behind the scenes of successful AI.

The reality is that many companies aren't ready. While organizations are making progress they aren't there yet. And the point is that a lot of leaders are trying to move forward, but they will only get so far, because they really don't get what is involved. So, I want to help them get it. This book examines the organizational components, data foundations, analytics foundations, skills, and governance needed to succeed with AI. It is geared toward directors and vice president of data and analytics as well as CEOs, CTOs, and other C-suite members who want to understand what is involved.

Many executives claim to support AI and analytics but lack a deep enough understanding of what it takes to build a truly successful program. It is not just about investing in cutting-edge technology; success depends on the right strategy, the right resources, the right tooling, and, most importantly, a cultural shift within the organization.

This book is designed to provide practical insights and real-world best practices drawn from my experiences in the field, my years of research, as well as the dozens of interviews I've done for this book. It is not a technical deep dive. Instead, it is a strategic guide that highlights the organizational and technological factors that lead to success with data, analytics, and AI. It is meant to discuss the landscape in a way that data and business professionals can understand, so they can understand what will be needed to succeed. It discusses the latest trends in data and AI, from data products to the data fabric, from data governance to responsible AI, and from building models to operationalizing them in production.

Throughout this book, you will find case studies and interviews with industry leaders, those who have successfully navigated the complexities of AI adoption. Their stories will provide valuable lessons and actionable strategies that you won't find in other books.

We are at a pivotal moment. AI is reshaping industries, redefining competitive advantage, and revolutionizing the way businesses operate. This book will help you navigate these changes with clarity, confidence, and a strategic mindset.

How This Book Is Structured

This book is divided into distinct sections that weave together to tell the story of what it takes to succeed with data and AI.

• Part I-The AI Imperative: Part I sets the stage for the rest of the book. It looks at what is happening in enterprises today, the promise of AI, how some organizations are not yet ready but being pressured to implement it, common mistakes, and the five key dimensions for AI readiness and maturity.
• Part II-Building the Data Foundation for AI: Part II examines the data foundation needed for AI. Many organizations try to jump ahead of this, especially with generative AI use cases; however, when they realize they need company data for their applications, they aren't prepared. The section discusses the new kinds of data organizations need to utilize (such as unstructured text data), the problem with data silos, and how organizations are trying to unify their data and create a modern architecture, and what is involved with this process. It also looks at other data management issues such as findability, useability, and data quality and trends in this area. It ends with newer trends in data management such as data products.
• Part III-Democratizing Analytics: Part III focuses on how implementing BI has often been a precursor to moving forward on the analytics journey. There are important lessons that companies can learn from implementing BI. Additionally, democratizing BI and making it available to more users serves numerous purposes; it frees up data analysts and other analytical roles to do more advanced work; it also starts to build the culture around analytics and gets people data literate, which will be important as they move into more complex analytics such as AI. This section describes why it is important and best practices for rolling it out into the company and getting buy-in. With the advent of generative BI, this becomes even more important.
• Part IV-The AI Build-out: Ecosystem, Models, Operations, and the Shift to Agents: Part IV examines best practices for building models and aligning them with business strategy. It also will look at the skills to build traditional and newer AI applications, a high-level overview of what is involved, as well as what's new with generative and agentic AI. In this section, the notion of using generative AI with company data is introduced, along with the challenges that entails. Here, I'll also talk about operationalizing AI applications and discuss what is involved from a technology and organizational perspective. This is about putting these models into production, the new roles, and processes this involves.
• Part...