Don Schuerman keynote at PegaWorld iNspire 2023: The Autonomous Enterprise: The Future of AI and Automation

Ever wonder what all this tech disruption really means for your business? Yeah, we do too! It’s what keeps us up at night and gets us up in the morning. And, we have some thoughts. Join Don Schuerman, Pega’s CTO, as he helps us make sense of the future, shares his insights on topics like the autonomous enterprise and generative AI (for example, Open AI’s ChatGPT), and how the nature of enterprise IT is evolving. We’ll look at the tea leaves to see what the future holds for technologists and business leaders alike – and get a sneak peek into Pega’s future as we reveal upcoming innovations and technological advancements.

Transcript:

My name is Don Schuerman. My pronouns are he, him. I love that I get to be CTO at Pega, not because that is Cooking Tasty Omelets, but because I like to think of it as Chief Translation Officer. Me and my team, we get to look at all the amazing tech that's happening out there in the world, the incredible things that Karim and his team are doing, and we get to make sure that all of you understand the incredible value that that can deliver to your business. But this is a two-way conversation. We also get to spend time listening to you and understanding your strategies, your needs, your goals, and make sure that all of those are translated back into Pega's roadmap and our strategy.

And hopefully, you've gotten to see some of that over the last couple of days. Maybe it was an amazing demo in the Innovation Hub or something that really popped in Karim's keynote, or maybe from one of your peers in a breakout, you heard an insight or a way of solving a problem that got you thinking about something you could bring home to your organization. And hopefully you've had a little fun. Hackathon last night was great. It was so much better than the Zoom hackathons I've been throwing for the last couple of years. None of you have actually shown up at them, probably because at the end of the day, we're not really here for the dance parties, although they're great. We're here to talk about how technology, intelligence, automation, the ability to collaborate and bring both the power of intelligence in the tech, but that organic intelligence that all of us carry around together, to drive real change for our business to build an organization that is continuously optimizing but always with humans in the loop.

This is all about operationalizing agility and that's what Alan really meant when he talked about the autonomous enterprise, bringing together intelligence and automation so that we can continuously optimize our businesses.So as Chief Translation Officer, I'm going to take you through three metaphors to explain where we're thinking about this. First, I'm going to talk about the path to an autonomous enterprise and I'm going to talk about our cars. I'm not going to tell you about what you need to get there, and I'm going to give you a recipe. And finally, I'm going to share why Pega is uniquely suited to help you on that journey. And I'm going to talk a little bit about the DNA that is in every single one of your workflow and decisioning applications, because we are living in a world where autonomy and intelligence is everywhere. It's in our cars, right? When we get behind the wheel and we have assistance to help us drive. It's in medicine.

I have a friend who is an eye surgeon and he was explaining to me the other day that when he does surgery, he doesn't actually do the surgery. His movements are translated by AI to an autonomous robot that operates at the submicron level to peel back the cornea of an eyeball. Excuse me, I think I just disgusted myself there. That's one of the many reasons why I'm not an actual doctor, doctor. And then finally, some of us have used stuff like ChatGPT, right? Which are these autonomous agents. As Dr. Walker showed us, they seem to have a mind of their own. They declare love to us. They can write poems about the wonders that may reach all ladies and gentlemen at PegaWorld, they can do our children's homework and we're seeing autonomy already show up in our business.

Many of you are doing this. Maybe you're using customer decision hub to have adaptive models that continuously learn from your customer interactions to make every next one better. Or maybe you've started looking at process AI and it's using your process data to continuously find new ways to optimize the process to make it better and smarter. Or maybe you've taken case management and business rules to guide your agents so that like John said, you reduce that training time. Maybe you're even using things like voice and messaging AI to direct and interact with your customers digitally, continuously learning and improving that experience.

In fact, we did a survey. We talked to about 600 of your peers and 88% of them said that in the next 10 years, they will either be or be well on their way towards an autonomous enterprise. 75% of them said that AI will play a major role in their business in the next five years. I have no idea what the other 25% are thinking, but only 33% of the people that we talked to said that they are ahead of their competition in harnessing the power of AI and automation. Now, if we limited the survey to just people in this room, I hope that that would be a much higher number.

But what that tells us is that unlike that ChatGPT volcano that erupted, autonomous isn't going to explode overnight. It's something that we're going to evolve continuously into. So think about your car. Despite all the talk of like we're going to have a self-driving car, completely self-driving, unless maybe you take a taxi in San Francisco, right? I still think there's some gaps preventing us from getting there. One is regulatory. If two self-driving cars crash, who's liable?

Some of the reasons are cultural. I, for one, like to drive sometimes. I don't know if I'm completely ready to give that up. And some of it is technical, right? We have a lot of individual components that are becoming pretty powerful, but those of us who have worked on enterprise software applications know the challenge isn't the individual pieces. It's stitching them all together in a way that acts consistently and predictably. So even though a fully self-driving car may still be years away, there's lots of autonomy in our car today.

When I sit down at the wheel of my car, I have cruise control that will dynamically adjust the speed of the car when I'm in traffic so I don't have to keep putting on the brake. I have park assist that will warn me if there are obstacles in my way and automatically apply the brakes if I don't ask fast enough. I have drive select that will look at road conditions and change how the transmission responds so that I feel a tighter contact with the car. I have navigation that not only can route me to my next destination, but can continuously monitor traffic conditions and suggest new routes as traffic conditions change. And if I plug my phone in, I've got this voice activated virtual assistant that can create a playlist of music that sounds like The Grateful Dead, much to the delight of me and to the hatred of my kids, right?

There are autonomous pieces in the car today, and I think that's very much how our path is going to look as businesses, right? Many of you are already well on this journey. You probably had processes and maybe you still have some processes that are not only manual, they're completely unmanaged. You have no way of tracking where the work is being done. Maybe some of your employees show up at the beginning of the day and you print out a spreadsheet that tells them what work they should be focused on today. And one of the first shifts that you made was to take that work to become managed so that at least there's a structure for how the work gets done. Maybe you're using prioritization so that people are actually working on the most important thing rather than cherry-picking.

And that structure then gives you the places to hook in the automation, maybe using business rules to automate an approval or calling out to a web service so that you can pull the data that a user needs, or maybe you're calling a bot because you've got to go get data from a system that doesn't have an API, one of those legacy systems that we all still struggle with. And as you automate more work, you're building data, you're building history about how that work has gotten done, and that then allows your work to become more intelligent, to become self-learning, to apply things like process AI or adaptive models in CDH that are continuously listening and learning and getting that feedback so that they can more accurately predict the right action for a customer or the right action to take during a process.

And that's what sets us up to become autonomous. Self-optimizing, but there's that big asterisk that Rob put up there, right? Always with a human in the loop, but driven by the goals that your business wants to achieve and the constraints under which you want to and have to operate, but optimizing towards that outcome. Now, to let all that happen, you need to have the right stuff in there. You need to have AI powered decision and workflow automation. It helps to have a structure like the layer cake so that you can organize all these assets and these templates, these pieces of a workflow and decision and a user interface, a runtime engine in a center of business architecture that can work across all the different channels and all the different systems that you have, support for the distributed cloud and containerization, but also a way to stitch that distributed world back together.

In other words, you need a recipe. And I yesterday got a little education on recipes. In order to make a Western omelet, your recipe needs two things. It needs the ingredients. So you need eggs, you need butter. Trick to omelets, you need a lot of butter, more butter than you think you should put in the pan. You need onions, peppers, ham, cheese, salt, pepper. But you can't be like a three-year-old and just throw all this stuff in a bowl and whisk it up and be, "Ta-da, and omelet." You actually have to put it together in the right way. You have to follow the instructions. And the same is true for the autonomous enterprise. You need the right ingredients, you need the right pieces of tech to make this happen, but you need to put it together in the right way. And in the technology world, we don't say instructions, we say architecture.

So what are the ingredients for an autonomous enterprise? Well, there's a lot of stuff, right? You need some AI powered decisioning and some workflow automation. Come to think of it, you probably should have case management. In fact, there's a lot. You're going to need low code and channel independence and you want to have generative AI, but you want it to be responsible and you want to have business rules to both govern the AI but also automate some of your decisions. And you're going to have a great user experience, but it needs to work across channels including mobile and web. You're going to need to integrate with your existing systems, but you also want to have work that can be distributed across different places in the business. You want to have AI to optimize your processes, but also optimize your next best actions for your customer.

Excuse me. You probably are going to have robotics because there's going to be some systems without some APIs. You want insights and reporting so you can see everything that's going on. You want reusability and a marketplace where you can get different components. You want to have process mining so you can figure out where your bottlenecks are. There's going to be unstructured stuff, so you're going to need NLP and chatbots and you want to automate emails. And your event driven architectures are a really big deal. So you need to be able to plug into events so you can operate in real time. You're going to have DevOps, right? Because you want to be able to deploy this stuff. But that means you've got to have auto testing, you got to have cloud infrastructure and containerization. And of course, you want to have auto documentation so you can go explain to a regulator what you're doing, right?

That's not a western omelet. That's like one of those recipes where you got to go to amazon.com to find Sichuan peppercorns. But the cool thing is, if you've walked around the Innovation Hub, saw Karim and Rob's keynotes, looked at the breakouts, you're finding a lot of those ingredients in the Pega Infinity platform, AI power decisioning, workflow automation, rules, process mining, the Constellation UI, the DX API, cloud, the work we're doing with Pega Gen AI. You've got this incredibly well stocked pantry for the autonomous enterprise, and we've got you the architecture so that you can put it together in the right way.

That layer cake structure that Alan talked about that organizes all the pieces that you need, so you have a comprehensive set of what you need to describe your decisions and your workflows, organized for reuse and variation. A center out architecture so that you can run this stuff in a way that works across all the different channels that you have, process fabric and a low-code app factory so that you can work in a distributed way, but connect up what people are doing and ensure that your developers are following best practices.

This is what we've built, the low-code platform for AI powered decisioning and workflow automation to do, to bring you towards this autonomous enterprise so that you have everything you need. And that architecture is also key to how we're going to be able to help you harness the power of generative AI because I think you have all of the pieces in place to get to truly self-optimizing customer engagement, truly self-optimizing processes, and gen AI is now just rocket fuel to send it into motion.

But to explain why that's so, I want to talk in my third metaphor about the DNA inside of all of your applications. And I want to talk about a book I've been reading called The Song of the Cell, highly recommended it by an actual doctor, doctor named Siddhartha Mukherjee. And he talks about how the cell is the building block of every living thing, from the tiniest little amoeba to the beautiful, complicated, contradictory, amazing individuals that all of us are as human beings, right? And core to all of that is DNA. It's the information structure that tells the cell how to behave, how to grow, how to change, how to communicate with other cells. It's this incredible computer program really that captures all the information for a tiny little cell to become an incredible, unique, special, beautiful human being just like you. "Aw, that's really sweet, Don. Sounds a good beat. Read. I'm going to slide that right underneath my murder mysteries." What does that have to do with the autonomous enterprise?

Well, how many of you have used App Studio? Show of hands, put them up. All right, keep them up. Look around. If your hand is not up, I have a post PegaWorld call to action for you, which is go use App Studio. Go sign up for a Business Architect's Essentials course. Go to community.pega.com. Get a free system that you can use. Go use App Studio because that's how you can understand how these building blocks come together. You can put your hands down, and if you use App Studio, you'll see that you are capturing all the aspects of your application, the stages and steps, the persona, the data, the integration, the languages in a structure, in a beautifully intuitive and really powerful environment. App Studio asks you questions, you fill it in, and it answers those questions for what you need to run a workflow application. And under the covers, inside of the layer cake that Alan talked about, you are actually writing the DNA of your application. You're filling in all the information that that workflow and decisioning application needs to run and grow and communicate and evolve.

And that it's that DNA, that structure that is sitting inside the layer cake that then allows us to generate from all of this a running application for you to use. And that DNA is really powerful also when we talk to generative AI, because generative AI has all of this information, but it's really unstructured. You got to ask it the right questions. You got to do the prompt engineering that Dr. Walker was talking about. But because we know the structure of the DNA of a workflow application and a decisioning application, we know exactly the questions to ask. We can automate the prompt engineering, which is this emerging science of asking gen AI the right questions. And when gen AI comes back with an answer, we can automate the response engineering. We can take the response and plug it back into that DNA of the workflow application, updating all the right pieces, the stages and steps and the data models and the personas and the channels so that you can generate a running application now from the inputs from gen AI.

But the cool thing is it's the same DNA that's behind App Studio. So if your users want to go in and change or review or override or modify what gen AI selected, it's not buried in code. It's right there in the same beautiful and powerful and intuitive interface that they need to use. And so we're really excited about what's in Infinity 23, but this structure, this architecture is going to allow us to do some really cool things in the future. And I want to take a moment and talk to you about something that's not in Infinity 23, but where I think the future of the autonomous enterprise is going.

And I think if you're going to build an autonomous enterprise, you need an AutoPilot. Now, not an AutoPilot that replaces a human right. I got on a plane to fly here to Vegas like many of you, and that plane had an AutoPilot, but I wouldn't have gotten on that plane if it didn't also have a human pilot. That plane, however, is a really big complicated piece of machinery and technology, and I probably wouldn't have gotten on that plane if it didn't also have an AutoPilot helping the human out. And that's what we want to provide.

So imagine AutoPilot as an agent that uses both all the power inside of Pega, but also gen AI to complete tasks and achieve goals for you. So an end user might, for example, ask AutoPilot to say, "Hey, I want you to sign an appraiser to assess the claim and summarize it on the report." And AutoPilot will actually go create a new case in Pega, lay out all the steps that it's going to do, and start notifying the appraiser, emailing the customer. The user, hand firmly on the steering wheel, can go look at the case, see everything the agent is doing on their behalf. And when the agent is done, it'll come back and present a summary of the report right to the user, helping them get their work done faster.

Or imagine an AutoPilot for your developers, where a developer could say something like, "I want you to integrate a web service into my claim system to get me the weather." An AutoPilot will say, "Hey, great. I found one at weather.gov. I'm going to create a rest rule." Hand firmly on the steering wheel, the user says, "Yes, go ahead and do that." An AutoPilot runs off in the background, builds the connectivity to the web service, builds the data page, which is the structure in Pega that will hold that data and integrates it back, and then comes back to the user and says, "Here's what I've done." So you can go look at it, change it, review it, override it.

But where this gets really powerful is if you imagine an AutoPilot for your business. Imagine being able to say, "I want to optimize my claims process to minimize time while maintaining MPS." And AutoPilot will say, "Great. I'm going to use process mining and process AI, and I'm going to find bottlenecks." And it runs for a while and looks and watches the data and comes back and says, "Hey, found a bottleneck in your process, could save you some money. I can fix it by putting in a process AI prediction to override an SLA. You want me to do it?" User says, "Yes," hand firmly on the steering wheel. And then AutoPilot goes and makes the change, routes it to a system architect to actually get promoted, and then continues to monitor the process for the next optimization opportunity.

We've been talking about self-optimizing processes and self-optimizing decisions for a long time, and I truly believe with the capabilities that we have in Pega, process mining, process AI, the case management, and adding to that, the power of generative AI, you can truly get to self-optimizing processes in your business. Again, none of this stuff is in Pega Infinity 23, but it's where we're going. And to help us get there, I need all of you to start taking advantage of some of the amazing stuff we've built today. Use App Studio, so all of you can raise your hands if we ask you again next year. Start deploying Constellation. Look at process mining and process AI. Take the time to invest in the customer decision hub and really change the way that you think about doing marketing, because just like autonomous cars, that will help you take the steps to set up for becoming a truly autonomous self-optimizing enterprise.

And I can't wait to come back at PegaWorld 2024, right here in the MGM Grand in Las Vegas, and tell you about all the exciting things that we've done. But I'm even more excited to come back and hear from you about all the amazing things that you've done with App Studio and CDH, and process AI and process mining and voice and messaging AI and Pega Gen AI. Because together we're going to find all more amazing cool uses for this technology, new ways to keep our humans in the loop while operationalizing agility. So I'm going to come back to some of the predictions that Alan made on day one.

In a year from now, you're going to have Pega expertise at your fingertips, prompting you right inside of App Studio with the help that a developer needs to do not just simple, but even more sophisticated tasks. Through things like Constellation and App Studio and Gen AI, we will have doubled developer productivity so that we can solve more problems at more scale than ever before. You will have a universe of solutions that have been developed by our partners on Pega Launchpad, and thanks to Process Fabric and the DX API, it'll be easy to just weave those solutions into the rest of your business.

And through all of that together, we will have continuously built organizations that can change, that can react, that can optimize. Volcanoes like GPT keep erupting. We don't know where the next change is going to come from, but I do know that together as this awesome community, both at the conference and on the dance floor, we can come together to use our organic intelligence and the power of all of this technology to truly build for change. Thank you so much. Enjoy the rest of PegaWorld, and we'll see you next year.