But this isn't just another cycle. Beneath the frantic activity, a tectonic shift is occurring, far more significant than the move to the cloud or the rise of agile development. The very essence of value in enterprise technology is changing. For three decades, the IT services industry has made its fortune by translating messy business processes into rigid, structured code. We built the unshakable Systems of Record.

Now, our clients are demanding something new. They don’t just want to record what happened; they want to understand why it happened and intelligently predict what will happen next. They need Systems of Reality.

This is not a simple transition from one technology to another. It's a fundamental change in our raw material—from the logic of code to the nuance of context. For IT services firms, this isn't an opportunity to add a new service line; it's a mandate to completely rewire our operating model.

Part 1: Our Legacy of Concrete Code

Let's rewind to 2010. Imagine the kickoff meeting for "Project Titan," a massive, multi-year ERP implementation for a global logistics company. The room is filled with dozens of our best and brightest: project managers, business analysts, and an army of developers. For the next three years, their lives will be consumed by translating the client's arcane shipping logic into millions of lines of bespoke software Every single exception had to be anticipated and hard-coded³. What happens if a container is held up in customs in Singapore? There’s an if-then-else statement for that. What if a snowstorm in Denver delays a shipment? A developer in Bangalore writes a specific function to handle it. The business process documents are measured in pounds, and the success of the project is measured in function points and on-time delivery of code modules.

This was the golden era of the System of Record. Our economic engine was application.  development, system integration, and the endless cycle of maintenance that followed⁴. We were masters of building digital concrete. We poured foundations of SAP, Oracle, and Salesforce, and they were strong, reliable, and utterly inflexible. If the business wanted to change a process, it required a formal change request, a new budget, and a team of developers to ceremoniously break open the concrete and re-pour it.

Services firms flourished by supplying the armies of coders this work required⁵. Our value was in our ability to meticulously translate human ambiguity into machine certainty. The systems we built were the definitive source of truth for what happened, but they were blissfully ignorant of the why. They were perfect recorders of a world that no longer exists.

Part 2: The Model is the New Operating System

The cloud era shifted the ground beneath our feet, but it didn't fundamentally change our work. It simply moved the concrete factory to someone else's data center. The focus changed to managing "workloads"⁶, but we were still, by and large, building the same rigid applications.

The AI era is different. It vaporizes the concrete. The foundational AI models we see today are the new "microprocessors"—the central engines around which everything else will be built⁷. But I’ll take that analogy a step further: for the enterprise, a large language model is the

New operating system.

Think about it. An OS manages resources (memory, compute) and provides common services to applications. A foundational model does the same for a much more valuable resource: understanding. It provides services like reasoning, summarization, translation, and inference. In this new world, the role of an IT services firm is no longer to write applications for the operating system. Our new role is to orchestrate reality into the operating system.

Let’s revisit our logistics client from Project Titan. Today, they don't need us to write a million lines of code to manage shipping exceptions. Instead, they need us to build a System of Reality. This system would be powered by a foundational model fed with a continuous, curated stream of context:

• Real-time shipping manifests and GPS data.

• Historical weather patterns and future forecasts.

• Port authority labor union negotiation statuses.

• Local news reports on traffic and road closures.

• Customer sentiment analysis from emails and service calls.

• The entire corpus of international shipping regulations.

The "application" is no longer a rigid program. It's a sophisticated, dynamic prompt that asks the model, "Given this specific shipment's destination, contents, and priority, what are the top five potential risks in the next 72 hours, and what is the optimal contingency plan for each?" The value isn't in the code that builds the user interface. The value is in the painstaking work of identifying, sourcing, cleaning, and piping all that context into the model. The raw material has changed, and so must the factory that processes it.

Part 3: Rise of the Context Engineer, Hero of the New Era

This paradigm shift demands a new hero. For the last twenty years, the star player in our industry was the "10x coder," the brilliant programmer who could solve complex algorithmic problems. That hero is now giving way to the Context Engineer.

The Context Engineer is a new breed of consultant, a hybrid thinker who is part business strategist, part data scientist, and part AI whisperer. Let’s imagine a day in the life of one. Her name is Maria, and she’s working with a large retail bank. Her mission isn't to "build an app"; it's to reduce customer churn.

• Monday: Maria doesn't open a code editor. She convenes a workshop with relationship managers and mortgage advisors to understand the subtle, human cues that signal a client might be unhappy. They talk about life events—a child going to college, a parent getting sick—that often precede a customer moving their accounts.

• Tuesday: She partners with data engineers to create a "context pipeline." They pull data from the bank's CRM, but they also ingest transcripts from call center interactions, sentiment from social media mentions, and even data from public records about property sales in a client's neighborhood.

• Wednesday: Maria spends the day crafting and refining prompts for the bank's foundational AI model. She's not coding; she's teaching. She’s showing the model how to connect a series of small, seemingly unrelated events (a large withdrawal, a customer service complaint, a login from a competitor's IP address) into a coherent narrative of churn risk.

• Thursday: She designs a RAG (Retrieval-Augmented Generation) system. When a relationship manager is about to call a high-value client, the system automatically pulls the latest context—market news that might affect their portfolio, recent company announcements, even notes from the last call—and feeds it to the model.

• Friday: Maria presents her work. She doesn't demo a user interface. She demos a new capability. She shows the bank's leadership a real-time dashboard that flags at-risk clients and, more importantly, provides the relationship manager with a set of contextually aware, empathetic talking points for their next conversation. She has built a System of Reality, a living model of the bank's customer relationships.

This is not science fiction. Companies are already building these new capabilities. In a landmark move, Cognizant and Workfabric AI announced a plan to deploy 1,000 Context Engineers, signaling a formal recognition of this critical new role. The race is on, not to hire the most coders, but to cultivate the most insightful context-builders.

Part 4: A Full-Scale Reboot to Escape the Strategy Industrial Complex

For too long, the IT services industry has been a key player in what I aptly named the "Strategy Industrial Complex." This is the decades-old model where elite consulting firms create beautiful PowerPoint strategies, which are then handed off to the IT services giants to be implemented over years, often losing the original intent in the translation to rigid code.

Strategy and execution were two separate, disconnected worlds. In the era of contextual computing, that model is obsolete. The strategy is the context you feed the model. A company's unique insight, its proprietary data, its understanding of its customers—that is its strategy. The act of building a System of Reality is the act of embedding that strategy into a live, breathing, intelligent entity.

You can't do this by simply adding an "AI Practice" to your existing organizational chart. It requires what I christened a "full scale reboot" of the IT services delivery model. It means breaking down the silos between strategy, data, and application development. It means organizing teams around business outcomes (like "customer churn" or "supply chain resilience"), not technology stacks.

The old model of selling a bank a "CRM implementation project" is dead. The new model is selling a "customer centricity capability." That capability is not a static piece of software. It's a dynamic system that continuously learns from new context, making the bank's strategy more intelligent every single day. The strategy is no longer a document that gathers dust; it's the living brain of the enterprise.

Writing the New RFP

This transition is daunting, but it is also the single greatest opportunity in the history of our industry. The firms that successfully navigate this shift will not be seen as vendors or implementers, but as true strategic partners, the curators of their clients' digital intelligence.

But it requires us to change the very way we define and sell our work. The RFPs we write and respond to must evolve.

The Old RFP Measured:

• Billable hours and staff headcount.

• Function points and lines of code.

• Server uptime and milestone completion.

The New RFP Must Measure:

• The quality and breadth of contextual data integrated.

• The accuracy and relevance of the insights generated.

• The speed to business outcome.

• The tangible impact on the metric that matters.

The future of IT services is not about being the world's best builder of digital factories. It's about being the world's most trusted guide in the digital jungle. Our clients are no longer buying code; they are buying insight. They are not buying Systems of Record; they are buying Systems of Reality. And the firms that understand this will not just survive—they will lead the next decade of technology.