If there was one topic that dominated conversations at PegaWorld this year, it was AI. AI agents. AI-powered self-service. Autonomous workflows. Intelligent automation.

And rightly so.

After a period of experimentation, the conversation is moving beyond whether AI can create value. We now have many examples showing that it can. Organizations are already using AI to improve productivity, accelerate service, increase employee effectiveness, and deliver better customer experience. But after spending the week listening to customers share their stories, one theme emerged again and again.

The organizations seeing the greatest impact aren't simply deploying more AI. They're connecting AI to the work that drives customer outcomes.

We don't have an AI problem. We have a work problem.

Customer service leaders today face a challenge that will sound familiar to almost everyone. How do you reduce operational costs while simultaneously improving customer experience? It's the challenge that sits behind almost every transformation initiative. AI appears to offer the perfect answer, which explains why investment is accelerating so rapidly across the industry.

Every organization is exploring how AI can improve interactions, deflect contacts, support employees, and drive efficiency. Yet despite all the excitement, many leaders are finding that the results don't always match the ambition.

Why? Because most transformation projects naturally start with the things everyone can see.

• The conversation.
• The channel.
• The agent desktop.
• The digital experience.

But the reality is that customer service outcomes are rarely determined by the conversation itself. They're determined by the work sitting behind it.

• The workflows.
• The systems.
• The processes.
• The knowledge.
• The data.
• The people.

Let’s be honest: customers don't care how intelligent an AI assistant sounds. They care whether their problem gets solved. And that is what several of the most compelling customer stories at PegaWorld reinforced.

Lesson one: Connected work creates better outcomes

One of the standout stories came from Blue Shield of California. Their challenge wasn't a lack of technology or employee expertise. It was fragmentation.

Nearly 1,900 service representatives were supporting members across more than 20 different systems. Every interaction required employees to navigate multiple applications, search different knowledge repositories, and manually piece together customer context. The issue wasn't that employees lacked information. The issue was that information, processes, and workflows were disconnected.

That's a crucial distinction. We often talk about customer experience transformation in terms of what customers see. But customers ultimately experience the consequences of how work happens behind the scenes. If employees are navigating multiple systems, customers feel that complexity through longer wait times, inconsistent answers, and repeated questions.

Rather than immediately focusing on AI, Blue Shield first focused on simplifying the work.

They brought together customer information, claims, eligibility, benefits, billing, workflows, and knowledge into a unified environment. The value wasn’t only operational efficiency. It also created a stronger foundation for everything that came next.

For me, this was one of the biggest lessons from PegaWorld. The organizations achieving the greatest value from AI aren't layering new technology on top of fragmented operations. They're creating connected foundations that allow AI to perform at its best.

Lesson two: Workflow is emerging as the backbone of enterprise AI

Another recurring theme throughout the event was orchestration. Because if fragmentation is the problem, orchestration is increasingly becoming part of the solution. One customer story that perfectly illustrated this came from enGen.

What fascinated me about their approach was that they didn't begin by asking, "Where can we add another AI assistant?" Instead, they asked a different question. "How do we coordinate work across increasingly complex customer journeys?" That's a fundamentally different perspective.

Their team focused on creating an orchestration layer capable of bringing together data, workflows, care teams, and decisions across complex healthcare journeys. In the example they shared, a single event - a customer being discharged from the hospital - triggered a coordinated chain of actions.

• Customer outreach was initiated.
• Work was assigned to the right people.
• Context was assembled automatically.
• AI generated insights and summaries.
• Care recommendations were refined.
• Next steps were coordinated.

What's important is that AI wasn't operating independently. It was operating within the workflow. And that's where things become truly powerful. Because customers don't experience isolated interactions. They experience journeys. The next phase of AI isn't simply about creating smarter conversations. It's about creating more connected outcomes. And that requires AI, people, systems, processes, and data working together as part of a coordinated operation.

Lesson three: AI needs structure to scale

A third lesson that came through loud and clear was the growing importance of governance. A year ago, much of the conversation centered on what AI could do. Today, organizations are increasingly focused on how AI can be operationalized safely, consistently, and predictably. That's a healthy evolution. AI is incredibly powerful, but power without structure rarely ends well.

For enterprises running millions of mission-critical processes, predictability is non-negotiable. Whether reviewing a fraud claim, handling a customer complaint, processing a loan application, or generating a regulatory disclosure, operations depend on consistency. Customers expect reliability. Regulators demand accountability. That's why workflow matters. It provides the structure, governance, and guardrails needed to ensure AI doesn't just act intelligently, it acts predictably, responsibly, and in alignment with business outcomes.

One of the most interesting examples came from Mizuho Bank. As one of Japan's largest financial institutions, Mizuho sees tremendous potential in AI. But they also operate in an environment where governance, compliance, and accuracy are non-negotiable. Their approach was simple but powerful. AI shouldn't operate outside the business process. It should operate inside it.

Workflow provides the structure, controls, transparency, governance, and human oversight that enterprises require. AI provides intelligence, recommendations, and automation. When those two capabilities come together, organizations can begin scaling AI confidently across mission-critical operations. This feels particularly relevant right now. The challenge for most enterprises is no longer proving that AI works. The challenge is embedding AI into the processes that matter most.

Lesson four: Trust is the real driver of adoption

The final story that stood out came from Sun Life. Like many large organizations, Sun Life didn’t begin its journey with ambitious plans for end-to-end automation. It started with a single business challenge. Knowledge.

Their challenge was helping employees access accurate knowledge faster across large volumes of content and complex information sources. Their response was to introduce Pega Knowledge Buddy (now named Pega Knowledge Agent™). But what interested me wasn't the technology itself. It was how they approached adoption.

They started small. They measured outcomes. They learned from the results. They refined the experience. Then they expanded.

Along the way, something interesting happened. The AI began exposing weaknesses that had always existed beneath the surface: incorrect articles, knowledge gaps, oversized content, and articles that needed refining. What initially appeared to be an AI initiative quickly became an operational improvement initiative. And in many ways, that's one of AI's most valuable attributes.

It shines a light on underlying inefficiencies that organizations may not have fully appreciated before. There's another lesson from Sun Life that I think every organization should take seriously.

Throughout their presentation, they repeatedly emphasized one word. Trust. Not just architecture. Not just prompts. Not just models. Trust. Trust that employees can rely on recommendations. Trust that AI is helping them perform better. Trust that the outcomes are predictable.

Ultimately, successful AI transformation isn't a technology journey. It's a trust-building journey. Technology enables it. Trust sustains it.

The real takeaway from PegaWorld

If I had to summarize my biggest takeaway from PegaWorld, it would be this: AI changes the interface. Connected work changes the outcome.

The organizations pulling ahead aren't necessarily the ones deploying the largest number of AI agents. They're the ones creating connected environments where AI, workflow, knowledge, data, and people work together as a single system.

They're moving beyond isolated pilots. They're moving beyond point solutions. And they're redesigning how work gets done. Because at the end of the day, customers don't care whether an outcome was delivered by an AI agent, a human employee, or an automated workflow. They care whether their problem was solved.

That's why I believe the future of customer service won't be defined by who has the most AI. It will be defined by who can most effectively combine AI with workflow, orchestration, governance, and trust.

The stories we heard from Blue Shield of California, enGen, Mizuho Bank, and Sun Life all reinforced the same point. AI is absolutely transformative. But AI alone isn't the finish line. The real opportunity lies in creating connected service operations where work flows seamlessly across people, systems, knowledge, and AI to deliver better outcomes for customers.

And judging by what we saw at PegaWorld, that future is already starting to take shape. Watch the replays here.