This article was based on the interview with Don Schuerman, CTO at Pega by Greg Kihlström, AI and MarTech keynote speaker for The Agile Brand with Greg Kihlström podcast. Listen to the original episode here:
In enterprise marketing, we live with a peculiar paradox. We are tasked with driving innovation, creating seamless customer journeys, and responding to market shifts with a nimbleness that borders on clairvoyance. Yet, we are often tethered to technology stacks that resemble a geological cross-section, with layers of legacy systems, undocumented processes, and business logic locked away in digital vaults no one has the key for. The brightest marketing strategies can, and often do, grind to a halt against the unyielding wall of technical debt. This is the quiet frustration of the modern marketing leader: the vision is clear, but the path is blocked by the ghosts of technology past.
So, when the conversation turns to Artificial Intelligence, it’s understandable that most of the excitement gravitates toward the new and the creative—generating ad copy, designing visuals, or powering the next generation of chatbots. But what if the most profound, most valuable application of AI in the enterprise isn’t about creating the new, but about understanding the old? What if AI’s greatest immediate contribution is the unglamorous but wildly expensive work of untangling decades of complexity? This is where the conversation shifts from the hype of generative AI to the pragmatic power of agentic AI. It’s a move away from seeing AI as merely a creative assistant and toward viewing it as a strategic partner in modernization, one that can finally bridge the chasm between a marketing leader’s vision and IT’s reality.
The Real Magic Happens at Design Time, Not Just Runtime
Much of the discourse around AI agents focuses on their “runtime” applications—autonomous agents executing tasks, managing customer interactions, or optimizing ad spend in real-time. While valuable, this overlooks a more foundational challenge. Before an agent can execute a flawless customer journey, that journey first has to be designed, built, and connected to the underlying systems. For many organizations, this is where projects stall. The process of translating a business need into a technical blueprint is fraught with misinterpretation, discovery workshops, and the painstaking archaeology of figuring out how the old systems even work.
Don Schuerman argues that the most significant immediate opportunity lies in applying AI agents to the design of these processes. By focusing AI’s reasoning and creative capabilities on the front end of the development lifecycle, organizations can achieve a powerful balance: leveraging the exploratory power of Large Language Models (LLMs) to design and modernize, while deploying more predictable, deterministic AI for the actual execution. This avoids the risk of a “creative” AI making unpredictable decisions with customer data or financial transactions in a live environment.
“So what we want to do is I want to use the reasoning power to design the right processes and design the right predictive models to then deploy those at runtime where I operate with a much higher degree of one predictability and consistency. But two, I actually operate faster and cheaper because I’m not calling the expensive large language models for every single interaction that I need to have.”
This is a critical distinction for any marketing leader. It means we can use AI to accelerate the path from idea to execution without introducing unnecessary risk into our customer-facing operations. It reframes AI not just as an executor, but as a co-architect. The expensive, unpredictable reasoning of an LLM is used once, during design, to create a stable, efficient, and cost-effective process that can then be executed millions of time at runtime. It’s the difference between brainstorming with a creative genius to write a script versus having that same genius ad-libbing live on air during a broadcast. One is a strategic use of creativity; the other is a recipe for brand-damaging chaos.
AI as an Archaeologist for Your Tech Stack
Every seasoned marketing leader has a story about the “one system” that can’t be touched—the mainframe, the AS/400, or the labyrinth of Lotus Notes applications that somehow still run a critical part of the business. These systems are often black boxes, their inner workings understood only by a handful of veterans nearing retirement. Attempting to modernize them is like trying to renovate a historic building with no blueprints; it’s slow, expensive, and you’re terrified of knocking down a load-bearing wall.
This is where agentic AI can function as a digital archaeologist. Instead of relying solely on human teams to spend months deciphering old code and interviewing subject matter experts, AI agents can be fed the raw materials of the past—source code, process documents, even screen recordings of users interacting with the old system—and reverse-engineer the business logic trapped within. This dramatically changes the economics of modernization. The initial, high-effort discovery phase that kills so many projects is automated, allowing human experts to focus on the more valuable work of reimagining and improving the process, not just documenting it.
“We’ve even had clients who’ve taken screen cams like Camtasia’s of people using mainframe systems and uploaded that into Blueprint and Blueprint will reinvent those business processes buried in the mainframe application. But in a couple minutes, you’re actually got a working prototype of a cloud-based system that’s modernized, that’s got agents, that automates things.”
For a marketing leader, this is transformative. The conversation with IT shifts from “We can’t change that journey because it touches the mainframe,” to “The AI has mapped the old process; how do we want to design the new, better one?” It turns a roadblock into a starting line. The ability to generate a working prototype of a modernized application in minutes, based on the digital artifacts of a legacy system, isn’t just an efficiency gain; it’s a fundamental change in how business and technology can collaborate to drive transformation. It makes the impossible, possible, and the prohibitively expensive, achievable.
From Creator to Conductor: Redefining the Marketer’s Role
Naturally, the rise of capable AI agents prompts questions about the future of marketing teams. If an AI can draft campaign strategies, create offers, and design workflows, what is left for the human marketer? The fear of obsolescence is real, but it often stems from a limited view of AI as a replacement rather than an amplifier. The most effective leaders are already reframing the conversation around scale and impact.
Schuerman shares a powerful anecdote about a copywriter who initially viewed generative AI as a threat to her craft. Over time, her perspective shifted. She realized that her expertise wasn’t just in writing a few perfect headlines, but in understanding the brand’s voice. By using AI, she could transition from being the sole creator of a limited set of assets to being the conductor and trainer of AI agents, ensuring that every piece of text generated across the entire organization—from sales emails to support articles to HR documents—was on-brand. Her impact went from dozens of assets to potentially millions.
“…now all of a sudden my role becomes a, I’m able to empower and scale what I personally can impact in significant ways. And I think getting people to understand that, you know, if you embrace these tools, they are scale tools. They allow you to increase your impact into areas you might not have thought before.”
This is the mental model marketing leaders must cultivate within their teams. Our value isn’t just in the direct execution of tasks, but in our strategic oversight, our creative judgment, and our deep understanding of the customer and the brand. AI agents become the instruments, but the human marketer is the conductor who brings it all together into a coherent symphony. The role evolves from doer to strategist, from individual contributor to force multiplier. It’s a shift that requires a willingness to let go of the old way of doing things, but the reward is a level of influence and scale that was previously unimaginable. We must remember, as Schuerman points out, that we lead people, but we configure software. Conflating the two is a path to poor leadership and ineffective technology implementation.
The path to a truly agile, customer-centric enterprise runs directly through the tangled jungle of its legacy technology. For years, the cost and complexity of clearing that path have been a significant barrier to progress. We’ve been told to “rip and replace,” a strategy that is as risky and disruptive as it sounds, or to simply build around the edges, leading to an even more complex and fragmented technology landscape. Agentic AI offers a third, more intelligent path: to understand, redesign, and evolve from the inside out.
For marketing leaders, this represents a fundamental opportunity. It’s a chance to move beyond the cycle of frustration where ambitious plans meet technical realities. By championing a pragmatic approach to AI—one focused on design-time modernization, legacy deconstruction, and the empowerment of our teams—we can become the true architects of our enterprise’s transformation. This isn’t about chasing the latest buzzword. It’s about wielding a powerful new set of tools to solve our oldest and most stubborn problems, finally freeing our organizations to deliver on the promise of modern marketing.
