New Book: The Intelligence Layer (How AI is Becoming a New Operating System for Business)

Why We Wrote This Book: The Intelligence Layer

When I teach machine learning, I often ask students a deceptively simple question: What do you actually do, once you’ve trained your model? Do you deploy it, tune it, scale it, forget about it, or build something bigger on top of it?

That question led me and a small group of collaborators to a much broader reflection: if AI is becoming a layer in every business and every product, how do we navigate that transition intelligently?

This book - The Intelligence Layer - is our answer. It’s not just about AI models or algorithms. It’s about the transformation of business thinking, decision-making, and competitive strategy in an era where intelligence is no longer just human.

The Seed of an Idea: Why Another AI Book?

Today, AI touches nearly everything: customer service, diagnostics, fraud detection, logistics, even art. And yet, when I talk to executives, product managers, or engineers, I still often hear a kind of quiet uncertainty: We know AI is important - but what exactly should we do with it?

This book grew out of a desire to answer that question - not at the technical level of loss functions and gradient descent, but at the strategic level where AI meets real-world complexity.

Alex and I set out to write a book that helps you do three things:

1. Understand how AI fits into your business architecture.

2. See clearly through the hype and focus on what creates real value.

3. Execute - not by hiring unicorns, but by thinking like one.

That pivot - from implementation to leverage - is where we believe the next frontier of AI transformation lies.

Why “The Intelligence Layer”?

Alex and I called the book The Intelligence Layer because increasingly, AI is not just a tool. It’s an abstraction layer that sits on top of business processes and turns data into direction.

Think of it as a new interface - not between humans and machines, but between intent and execution.

In software, we used to write instructions line by line. Now we describe outcomes, and AI systems figure out the steps. In business, we used to make decisions based on reports. Now AI systems can suggest the decisions themselves - and sometimes even take them.

The implications are profound. Just as cloud computing abstracted away infrastructure, AI is abstracting away the manual logic of business operations.

But this shift comes with challenges. When intelligence becomes infrastructure, how do you debug it? Govern it? Trust it? These are questions we explore in the second half of the book.

How We Wrote It

Writing this book was an iterative process - not unlike building a machine learning system.

We started by identifying the “target function”: What kinds of decisions do readers need help making? What mental models lead to good outcomes? What common mistakes can we help them avoid?

Then we collected “training data”: real stories from teams applying AI - successfully and unsuccessfully - across healthcare, finance, e-commerce, and logistics. We looked not just at technology choices, but at team structures, cultural dynamics, regulatory friction, and strategic trade-offs.

From there, we built out a kind of conceptual pipeline:

1. Foundations: What is AI, really, and what can it do now?

2. Business Architecture: How AI changes workflows, roles, and incentives.

3. Strategic Leverage: Where intelligence creates margin, speed, or differentiation.

4. Implementation: From data readiness to responsible deployment.

5. The Road Ahead: Navigating what’s coming - ethics, autonomy, and hybrid intelligence.

Each chapter went through many iterations. We asked ourselves, over and over: Is this helping the reader think more clearly? Could they apply this tomorrow?

If the answer wasn’t yes, Alex and I rewrote it.

A Few Core Ideas

If I had to summarize the book in five sentences, it might go like this:

1. AI is not a goal - it’s a lens. It lets you ask better questions about what your product, process, or team is really doing.

2. The first wins are almost always operational. Not visionary moonshots, but automating routine analysis, triage, and prioritization.

3. Every AI system is an opinion, embedded in code. You must design it like you would design a team: with clarity, accountability, and purpose.

4. Data is not just an asset - it’s an interface. Clean pipelines and labeled signals aren’t just for accuracy - they're for agility.

5. AI doesn’t replace people - it rearranges expertise. Great organizations won’t remove humans from the loop - they’ll redefine the loop.

Lessons From Building With AI (and Failing)

One of the privileges of working in AI is that I’ve seen many experiments - some that succeeded, and many that didn’t.

Some of the lessons we share in the book came from watching promising pilots fail because they were scoped too broadly, with too little iteration. Others came from small, focused efforts - like improving warehouse pick accuracy by 2% - that ended up saving millions.

We’ve also seen companies invest heavily in models, only to be blocked by lack of clear ownership for inference-time decisions. Or deploy a chatbot without anticipating the downstream escalation paths. Or build powerful AI tools that no one used - because the interface was wrong.

In almost every case, the problem wasn’t the model. It was the surrounding system.

What We Hope You Take Away

Ultimately, this is not a book about AI.

It’s a book about thinking clearly in a world where decisions, actions, and interactions are increasingly intermediated by machine intelligence.

It’s for product leaders trying to figure out where to apply AI in the user journey. For operations managers wondering how to prioritize intelligent automation. For startup founders looking to differentiate. For C-suite executives navigating where to build vs. buy. And yes - for engineers, too, especially those stepping into product or leadership roles.

We hope the book gives you:

• A vocabulary for talking about AI in a business-relevant way.

• A framework for identifying real opportunities (not just tech demos).

• A toolkit for leading AI-driven transformation, team-first.

A Closing Thought

The journey of AI has always been more than technological. It’s epistemological. It’s about how we learn, decide, and act.

Our goal in writing The Intelligence Layer was to contribute to that journey - not by predicting AGI, but by helping leaders make smarter choices today.

If this book helps you ask better questions, challenge default assumptions, and build systems that are not just intelligent, but wise - then we’ll have succeeded.

Thank you for reading. We’re excited to see what you build next.

 -  Denis and Alex