What if AI could design and execute your business processes for you? 🤯
The Futurum Group’s Keith Kirkpatrick is joined by Pega‘s Don Schuerman, CTO and Vice President of Product Strategy and Marketing, to discuss using Agentic AI responsibly to drive impactful business outcomes. Agentic AI is gaining momentum in enterprises, and Pega’s strategic approach is playing a critical role in modernizing legacy systems for transformative success! Find out how the company is empowering businesses to move beyond basic automation and embrace a new era of intelligent workflows.
Key takeaways include:
🔹From Reasoning to Design: Pega is empowering agents to design workflows, moving beyond runtime decision-making to create predictable and governed processes.
🔹Human-Guided Automation: This isn’t about replacing humans; it’s about augmenting them. Pega’s approach emphasizes human oversight and accountability, ensuring AI-driven workflows align with business goals.
🔹Orchestrating the Autonomous Enterprise: Pega is building the infrastructure for a future where agents seamlessly orchestrate complex tasks, driving continuous self-optimization across the entire business.
🔹Real-World Impact: From fraud detection to legacy system modernization, Pega is building practical solutions that streamline operations, enhance customer experience, and unlock significant value for enterprises.
Sponsored by Pega.
Learn more at Pega and about what’s next for systems transformations at PegaWorld 2025.
Watch the full video at Six Five Media, and be sure to subscribe to our YouTube channel, so you never miss an episode.
Or listen to the audio here:
Disclaimer: Six Five Media is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded, and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors, and we ask that you do not treat us as such.
Transcript:
Keith Kirkpatrick: Hi, I’m Keith Kirkpatrick, research director with the Futurum Group, and welcome to the Six Five Media Virtual Webcast. Today, I’m pleased to have Pega’s CTO Don Schuerman join me today to talk about the company’s approach to using Agentic AI responsibly to drive real business outcomes. We’re going to talk about how Agentic AI can be used to automate processes, improve customer experiences, and modernize legacy systems leveraging Pega’s Gen AI Blueprint. We’ll discuss how the Pega Gen AI Blueprint is central to agentic transformation and provide a preview of what attendees might be able to expect at this year’s PegaWorld. So, thanks for joining us, Don.
Don Schuerman: Great to be here.
Keith Kirkpatrick: Well, you know, I don’t think we can escape the fact that agentic AI is on the minds and lips of everybody in the software industry, really kind of gaining momentum across the enterprise. What’s Pega’s take on how to approach it and how to get it right?
Don Schuerman: Yeah, I think everybody is excited about the power of Agentic AI. When I talk to leaders at enterprises, I think there’s concerns about how do we leverage this power, but in a way that we have governance control. We don’t get thousands of agents kind of running amok and that we have to manage and keep an eyeball on. Right. And so what we’ve tried to get people to think about is, well, how do you think about agents as things in your business that can design, automate, and optimize the workflows that run your business? Because I think the business naturally knows how to think in workflows. They have governance structures that exist around workflows. And so the first part for us is how do we use agents to design workflows? And, you know, there’s lots of excitement about agent reasoning and agents being able to figure out what they want to do. And I think that’s great and really, really powerful. I think for the enterprise, you get a lot of that benefit with a lot less risk if you shift that reasoning to design time. So instead of the agent figuring out a runtime, hey, here’s how I’m going to handle this fraud that just showed up. Instead, I use an agent at design time to say, let’s define the best practice using all your great agentic reasoning and your ability to think slightly differently than how we normally do things to lay out what the new modern workflow and process for how we handle fraud is. And maybe I collaborate with the design agent, what we call Pega Blueprint. That’s really what Pega Blueprint is as it works as a design agent. Collaborate with humans and you get to here’s the design of our process and then you move that into, into an actual live system. But now you can ensure every request that comes in, you’re going to follow the laid out and agreed to best practice. Right. So you get that predictability that becomes really important. So that’s agents designing a workflow for you.
Keith Kirkpatrick: Right. Is that, you know, it sounds like that is the step that sometimes enterprises forget about doing that work before the work in order to make sure that Generative AI doesn’t sort of go off the rails, which is sort of a major concern for preventing.
Don Schuerman: Yeah, I mean, I think the challenge is, and part of it is if the agents that you’ve got at runtime are operating solely on text based prompts. Right. And again, I think it’s useful to put a human lens on this. If I wrote a paragraph of text and I gave it to 10 different humans, they might interpret it differently as to what they’re supposed to do. Right. And the same thing is true for agents. We know. Like if anybody has played with Chat GPT the same question three times, you get three different answers. Right. That’s actually a good thing if you’re trying to brainstorm. It’s a good thing if you’re trying to get design time ideas on how we can make this workflow better. It’s a bad thing if I’m trying to process fraud at a bank. I need to do that in a consistent way every time. So what we would say is rather than using text based prompts to define how agent operates at runtime, use the text based prompt to design the workflow with Blueprint and then feed the workflow, which now has the defined structure to it, into the agent at runtime. So the agent knows, I can maybe tinker around the edges, but I gotta hit all of these steps. Always.
Keith Kirkpatrick: Right. So you’re really kind of leveraging the power of agentic AI but doing it in a way where you can provide consistent and really predictable results in terms of workflows as opposed to, you know just–
Don Schuerman: Whatever the agent may come up with. Right. And then once you design the workflows well now you want, you want to be able to execute them agentically. So what does that mean? Well, that means one, I need to be able to actually have conversations with these workflows right. So what we’ve done is we’ve built the ability for you to define what we call an agent experience around the workflow. So you might say my customer service workflows. And I want those to be available to my customers via our website for them to use. So I’m going to take a collection of these workflows, I’m going to make them available to my customers. And what I’m going to use the large language model for is essentially as, without getting too technical, a semantic agent. Right. So that if a customer comes in, a customer might not know that there is a workflow called fraud dispute resolution.
Keith Kirkpatrick: Right.
Don Schuerman: Customers might just come in and say, hey, there’s a weird charge on my card that I don’t recognize. So we’re going to use the semantic agent to say, ah, weird charge you don’t recognize. That means we should run this fraud dispute resolution workflow.
Keith Kirkpatrick: Right.
Don Schuerman: Now I pick that up and because the workflow is defined, I can now, in a chat conversation, start collecting from the customer all the information we need. Well, what’s the charge that you’re worried about? Was it this one? Who, why, why is it unfamiliar to you? When did you figure out that this, when did you first notice, like all the information we need to know, need. But the agent is getting that information from the workflow. The workflow is sort of acting as the prompt to guide the agent to here’s what you should do next. Here’s what you should do next. So that uses the power of the large language model, but in a way that feels very safe and repeatable as opposed to unpredictable. And I don’t know where this is going to go.
Keith Kirkpatrick: Right. That makes sense. Now Don, you’ve mentioned just in your example about looking at fraud detection, are there other sorts of opportunities within the enterprise to really have agentic AI to make an impact in terms of real world business transformation?
Don Schuerman: Yeah. One of the biggest ones to me is this area of legacy transformation. Right. So every CIO I talk to has got mountains of technical debt. There is far too much of it. Budgets that are spent on sort of maintenance as opposed to innovation. Right. I’m just keeping the lights running on an old system that all the developers have retired for. And nobody’s afraid to touch anything because if I pull this thread, the whole sweater unravels. Right. And that’s always been a hard issue to tackle. And even when we’ve tackled it, a lot of what we’ve ended up doing with these legacy systems have kind of been like lift and shift. Like take our legacy system and we’ll move it to the cloud. Like I’ve literally seen, cloud migrations where we’ve migrated a system to the cloud and the new cloud based system is still a green screen. Like it’s still the same green screen interface, but it’s just running in a cloud emulator and it’s. That doesn’t change my business.But the power of agentic AI is. Well, maybe I could just take a video of somebody using my legacy application, just literally take a video of them stepping through it, describing what the steps are and what they fill in here and what they fill in there. And then maybe I could take some documentation of the database that’s the back end of that application. So what are the data structures and the DDL behind it? Feed that into the agentic model. So that’s going to be stuff that we’re going to showcase at Pega World, being able to upload this documentation into Blueprint. And Blueprint now lays out a modern version of what that workflow could be.
Keith Kirkpatrick: Yeah, Don, that’s really interesting because it’s sort of taking the idea that instead of having to, let’s say hard code something, you’re actually this. It sounds like the system is actually kind of being stepped through. What were the processes, you know, from a legacy system and then how can that be translated to a more modern one? Is that kind of what.
Don Schuerman: You’re exactly right. And the other thing is not only am I modernizing the systems, I’m actually modernizing the processes as well. Right. Because the other thing you want to be able to do as you go through this process, and especially if you can make this fast enough, is I don’t want to take my broken process from an old system and just put it on the cloud. I want to actually think through how that process could be made better. And the great thing that we can do with Blueprint is we use the agents to read and understand what the old process is, maybe pull in some information around what is the current best practice for driving that kind of process. Throw in some thinking about, well, boy, this process now needs to be able to support customers in a self service mode where they’re coming in through a chatbot and through a mobile app. Right. Now let’s find what that process looks like. And my guess is it’s going to have some of the same components, but the overall flow might be very different. Right. So I can accelerate that path from my legacy application to a modern application. And I’m using a genius AI basically to do that for me, I think that’s a really, really powerful use case because imagine the amount of IT spend and maintenance spend that frees up and, and the amount of innovation that could unlock at the enterprise.
Keith Kirkpatrick: Absolutely. And you touched on something that I think is worth going back to, which is the ability to go through and look at the processes that are being used to find out if there is a more efficient way of doing something or connecting systems that perhaps in the past didn’t need to be. Is that something that organizations need to really do in terms of really doing a modernization as opposed to, as you mentioned before, just a lift and shift to the cloud?
Don Schuerman: Yeah. And I think sometimes it’s really hard for enterprises to find those. Right. You know that one of the things, a CIO said this to me many years ago and it stuck with me because I think it’s so true. And that is when people live on the cow path, their natural tendency is going to be to pave it. Right, so you end up with a lot of. I think you’ve probably been through, I’ve been through programs that you do a lot of work to find out, like at the end, all I’ve done is basically repave the cow path. I’ve taken this broken process. The great thing is if I bring AI in and I use the kind of creativity and frankly, unpredictability of AI models for good, what AI can do, AI can go, hey, have you thought of looking at the process this way?
Keith Kirkpatrick: Right.
Don Schuerman: Have you thought of maybe, maybe you don’t need these five steps? Maybe you can just have that all be automated. And AI almost operates like a design thinking consultant coming to your business and saying, I can help you, even if it’s not your natural tendency, Mr. Biz Ops owner, I can help you step back and look at this process with fresh eyes. Right. And that’s really, really powerful because that helps solve the “people” challenge in a lot of this, which is how do you get people who have always thought of doing the work this way to think about doing it in a new way?
Keith Kirkpatrick: Right. Isn’t that sort of the, you know, the biggest challenge with any kind of modernization is the people component? I mean, if you think of, you know, people have spent 10, 20, 30 years doing things or thinking about things in a very specific way, and now we have technology that allows us to really take a step back in a more agnostic or unbiased way of looking at things for what they are, and then hopefully we find new, more efficient pathways. And it sounds like that’s what Blueprint is really sort of helping to do.
Don Schuerman: Yeah, I mean, that’s exactly, that’s exactly it. Right. Blueprint is a great way of just kind of shocking the system a little bit. And again, what you’re doing is you’re taking the fact that large language models and agentic AI have some degree of unpredictability to them, but you’re actually using that as a good thing. You’re using it like it’s great because it’s now going to generate ideas and it’s going to keep throwing new ideas at you, but then ultimately it’s still going to be dependent on a business person to go, this is what we’re actually going to do. Right. Because ultimately somebody in the business has got accountability for making sure that we process every customer fraud request the same way, and we process it according to best practices and guidance. So it starts to use Agentic AI truly as a creative partner business. And I think that’s just a really frankly underused but powerful capability that AI is bringing to us. And people like to talk about AI helping actual people be better, do their jobs better. This is a great way where you’re combining what AI can do and what humans can bring to get to something that is really powerful and maybe unique to both.
Keith Kirkpatrick: Right, right. I know that obviously Blueprint is going to be a really major focus at PegaWord this year. I was wondering if you might be able to give us a little sneak peek about what to expect and what you’re going to be speaking about during your session at the event.
Don Schuerman: Yeah. So I’m, I’m starting to talk about what I’ve been calling the age of Agentic workflows. Okay, so what does it happen? You know, we’ve been talking now for many years about this view of autonomous enterprise. Right. So how do we get continuous self optimization into the business? How do we design our businesses around the work that needs to be done and the outcomes that need to be delivered, and then continuously optimize everything in the business to deliver those outcomes more effectively. Right. You know, just in the same way that you put a destination into your car and your car helps you figure out how to get there as quickly and as safely as possible. Right. Same idea for the autonomous business. Well, I think core to that is this idea of an Agentic workflow and an Agentic workflow to me means it was designed by agents using tools like Pega Blueprint. But interestingly enough, we are also going to be showcasing some capability where maybe, maybe I get midway through this fraud request. And as I’m a fraud processor, I’m working on it and I’ve realized that this is a new type of fraud. I haven’t seen it before. Right. Well, maybe I want to in real time design a new workflow for myself. We’re going to show the ability to kind of take the Blueprint engine, but call it at runtime and design a new Ad Hoc process that I want to run but still have it under workflow control. So I asked the agent, how can I tackle this? And the agent could say, well, you should do this and this and this, and I can automate this for you. And then a human says, great, I agree, that’s the right path. I sign off on it now, let’s go do it and we execute it for you automatically. That’s really powerful.
Keith Kirkpatrick: Absolutely. Yeah.
Don Schuerman: And then we’re going to be talking about how you can plug these Agentic workflows into real time conversations with customers, real time conversations with employees, and also how these agentic workflows really start doing. Most of what they’re doing is orchestrating other agents to get the work done. Right. Because I think the other thing that enterprises really want is how do we maximize the efficiency and there is increased opportunity to say, great, well, this step doesn’t need to be manual. I can have an agent process that document for me. I can have an agent go do that research for me. Right. So how do we do that but do it under the umbrella of a workflow that is guided and governed. So again, to me, the key thing that we’re trying to do is how do I blend the power of agents with the predictability of a workflow? Because that’s really where I make this stuff work for real, at the enterprise.
Keith Kirkpatrick: Right. And it sounds like the, the other sort of critical component there is the fact that you do have humans who are serving as guardrails there. Because I think that’s critical to deploying Agentic AI within the enterprise.
Don Schuerman: It’s critical from a human perspective to service guardrails. It’s also critical because the way our whole society, legal structure, organizational structure, functions is I can’t assign accountability to an AI.
Keith Kirkpatrick: Right.
Don Schuerman: I can only assign accountability to a human. So how do I actually put those places where the human still is accountable, even though AI is doing large parts of the thinking and large parts of the actual work.
Keith Kirkpatrick: Yeah, that makes sense because at least at this point in time, that’s where we are.
Don Schuerman: At this point in time that’s where we are. And you know, again, given the state of our politics, I don’t know if we’re going to make sweeping changes to our legal system anytime soon. So.
Keith Kirkpatrick: I think that’s a discussion for another day..
Don Schuerman: For another day. Absolutely.
Keith Kirkpatrick: Well, I really appreciate you taking the time to speak with me today, Don. I think I’m really looking forward to all of the new things coming out of PegaWorld. I know certainly last year I had a lot of fun digging into Blueprint and playing around with it. But, you know, hopefully over the next several months, we’ll continue to see more innovation from not only you guys, but of course, your customers. So thanks again for joining me here.
Don Schuerman: It’s great to be here. And again, I would remind everybody that, like, if you’re curious, Pega.com/Blueprint anybody can go try it out. Anyone can kind of experience this stuff right now. So by all means, check that out. And I look forward to seeing you, Keith, and many others at PegaWorld.
Keith Kirkpatrick: Sounds great. Well, thanks again, Don, for joining me today here on the Six Five Media Virtual Webcast. Be sure to like and subscribe and we’ll see you again really soon.
Author Information
Keith has over 25 years of experience in research, marketing, and consulting-based fields.
He has authored in-depth reports and market forecast studies covering artificial intelligence, biometrics, data analytics, robotics, high performance computing, and quantum computing, with a specific focus on the use of these technologies within large enterprise organizations and SMBs. He has also established strong working relationships with the international technology vendor community and is a frequent speaker at industry conferences and events.
In his career as a financial and technology journalist he has written for national and trade publications, including BusinessWeek, CNBC.com, Investment Dealers’ Digest, The Red Herring, The Communications of the ACM, and Mobile Computing & Communications, among others.
He is a member of the Association of Independent Information Professionals (AIIP).
Keith holds dual Bachelor of Arts degrees in Magazine Journalism and Sociology from Syracuse University.
