Skip to main content

We'd prefer it if you saw us at our best.

Pega.com is not optimized for Internet Explorer. For the optimal experience, please use:

Close Deprecation Notice

Video

Kerim Akgonul keynote at PegaWorld iNspire 2023: The Times They Are AI-changing...

The world is constantly changing! And guess what? It’s changing again. How do transformative shifts like generative AI and analytics impact how you build low-code applications, engage with your customers, and provide differentiated service across channels? Join Kerim Akgonul, Pega’s Chief Product Officer, as he explores the latest Pega Infinity™ innovations in AI, customer engagement, and workflow automation so you can be ready for anything that happens – today and in the future.


Transcript:

Kerim Akgonul:
Hello, everyone. Thank you so much for being here. It is so nice to see you all here back in Vegas, thanks for making the trip. I feel like it's been so long. I can't believe it's been four years since we had this event live. Do you even remember what 2019 BC was like? No, no not that 2019 BC, 2019 before COVID. We were all here, we were at PegaWorld 2019 we were taking lots of selfies just like the party last night. It was great, we had a blast. And we're going to have a blast again this year. But I look back, it was such a simpler time. We were so innocent. We didn't know what was coming.

And then boom, all of a sudden we experienced a major shift in our lives. COVID hit and everything changed. Our work lives changed, our lives at home changed, the experience of our kids at school completely changed. It changed so much. We lost track of time. These days you bump into a friend that you haven't seen in a while and it's like, "Did I see you last summer? Or was it three years ago?" We don't know. We lost track of time, we lost months and years. It had a major impact.

But some things didn't change. The corporate objectives didn't really change that much. Enterprises are still driving towards digital transformation and going through major cloud migration so that we can provide amazing engagement to our customers. So that we can provide a seamless service experience to our customers across every single channel that they use. So that we can actually drive operational efficiency across the enterprise. And you know operational efficiency is really a secret word for doing a whole lot more work with less money which is the name of the game these days.

But these themes, we're familiar with these themes, we know how to work with them, we understand how to make progress towards these. But then just as we were walking out of our caves and our basement and ready to get back to work, boom, again, another tectonic shift hit us, and this time in tech. It feels like last November a volcano erupted with ChatGPT, and since then it's been GenAI everywhere and it's making a big change already. It's going to change how we work, it's going to change titles, skills, how we get work done. Forget about working from home, forget about the great resignation, forget about baking banana bread. When was the last time you heard about quiet quitting? Those aren't the headlines anymore, these are.

It's all about the generative AI, it's all about what we can do with generative AI. We can all see that it has so much potential, but we also see that there are a lot of questions with it. Can you trust it? Is it safe? Is there risk? Yeah. You can't you just ignore it and wait until it goes away. Nope, it's not going anywhere. And as you can see, Pega has been hard at work for years actually to bring AI into the enterprise. And we're really excited to be able to bring the value of generative AI in a safe and secure way to our customers in the next version of our product, Pega Infinity '23. Oh, yes, yes.

Now, Pega Infinity '23 has a tremendous number of great capabilities that are all intended to bring value directly to our customers and our partners. There's great new UI capabilities with Constellation all the way across the built-in accessibility capabilities. We have our (DX) API so mature that you can actually experience Pega applications across all the different channels. All the way across to a beautiful new robotic studio to help you build your applications. You heard about Process Fabric capabilities, you'll see a lot about process AI during this conference.

But today, in this presentation, I'm going to focus on, did you guess it, GenAI capabilities of the latest release. To do that we're going to actually show you the stuff by playing a game, a game show, that's never been played before. And we actually picked two names at random, and two people are going to come on stage to help us show you the latest capabilities. Please help me welcome our Head of Product Operations, Kara Manton, and our Chief Technology Officer, Don Schuerman, on stage with me.

Kara Manton:
All right.

Kerim Akgonul:
All right.

Kara Manton:
Hey, good luck, Don.

Don Schuerman:
Good luck, I don't need luck.

Kara Manton:
Okay.

Kerim Akgonul:
All right. Kara, Don, thank you so much for joining me on stage. We want to show what GenAI looks like in Pega Infinity '23 to our audience and we're going to play a game, a game that has never ever been played before.

Don Schuerman:
Never before.

Kerim Akgonul:
Yes. And the name of the game is Wheel of Pega GenAI.

Kara Manton:
Woo-hoo. Let's do it.

Kerim Akgonul:
All right. So this is how the game works. We have two contestants. I'm going to press the button and the wheel is going to go through a set of random process definitions that is going to be built from scratch, and then it's going to select a language that the end users need to experience it in. Here's how the game works. Actually, let's just select the process and the language first.

Don Schuerman:
Yeah, okay.

Kerim Akgonul:
You ready?

Kara Manton:
Sure.

Kerim Akgonul:
All right, let's go.

Kara Manton:
Let's see.

Kerim Akgonul:
Oh.

Kara Manton:
Okay.

Kerim Akgonul:
That's interesting. Here's how it's going to work.

Don Schuerman:
Okay.

Kerim Akgonul:
I am Turkish but I know you guys don't know Turkish so this should be interesting.

Don Schuerman:
I don't.

Kara Manton:
Okay.

Kerim Akgonul:
So, Kara, you're going to go to your workstation, okay, and you're going to use Pega Infinity '23 and build a loan application process, on Infinity '23, leveraging GenAI for Turkish speaker users.

Kara Manton:
Yeah, I don't speak Turkish.

Kerim Akgonul:
That's all right GenAI will help you.

Kara Manton:
Okay. Okay, great.

Kerim Akgonul:
And you're going to do this faster than Don Schuerman can make me a Western omelet, please.

Kara Manton:
Perfect.

Kerim Akgonul:
I'm starving. All right. Let's do it.

Kara Manton:
I think I have a good shot.

Kerim Akgonul:
All right. Get to your stations. Wait for my signal to go.

Don Schuerman:
Let me get my apron on, Kerim.

Kerim Akgonul:
All right. What does it say?

Don Schuerman:
CTO, cooking tasty omelettes.

Kerim Akgonul:
It's a great title.

Kara Manton:
This is perfect.

Kerim Akgonul:
All right. Ready, set, go. All right. We've got Kara's laptop on the screen, Don is getting ready. So what we're going to see is essentially Kara leveraging Infinity '23, she's going to build a process efficient from scratch. And you'll see, in the demo, every step of the way as she's defining the process and the rest of the application she'll be guided by the generative AI capabilities to make her more effective.

Don Schuerman:
Oh, oh, oh.

Kerim Akgonul:
Don, what are you doing? What is that?

Don Schuerman:
I'm putting on my onion goggles, Kerim, so I don't tear up on stage.

Kerim Akgonul:
Oh, we don't want to see you crying, especially when you lose. All right. Don is chopping some onions over there. And Kara, what are you up to?

Kara Manton:
Yeah, I'm just filling this form, I want to test my app. It's already working so Don might need to go a little faster.

Kerim Akgonul:
All right. Kara already got the first screens up and running. She's using some sample data creation capabilities, she's got some data model. Don is-

Don Schuerman:
Chopping some peppers.

Kerim Akgonul:
Chopping peppers for the Western omelet is a critical ingredient. Kara, she's getting into the data definitions, she's doing some live data stuff.

Kara Manton:
Yeah, I need a quick integration I think.

Kerim Akgonul:
We're going to get some integration built leveraging the GenAI capabilities to make her much more effective. It smells good.

Don Schuerman:
We're getting there, we're getting there.

Kerim Akgonul:
All right, all right. You can't make an omelet without cracking some eggs. All right. He's going really fast just so you know.

Kara Manton:
No pressure I guess.

Don Schuerman:
I got some shell in there, Kerim.

Kerim Akgonul:
All right. Kara, what do we have here?

Kara Manton:
I mean-

Kerim Akgonul:
All right, we have the application-

Kara Manton:
I have pretty much a working app here.

Kerim Akgonul:
We have the screen. Okay. You can build out the screens, you can straight-create the initial forms.

Kara Manton:
I mean, I think I'm done, Kerim.

Kerim Akgonul:
Wait, so what is Don doing though?

Don Schuerman:
I'm whisking the eggs, Kerim.

Kerim Akgonul:
Okay, good. Listen, Kara says she's done. That's not Turkish.

Kara Manton:
I was really hoping you wouldn't notice.

Kerim Akgonul:
No, translate it into Turkish. Jesus, go for it. All right, Don, you have plenty of time. She has to learn Turkish, go and get the ... Externalize all the phrases. She's going to get someone to translate for her.

Kara Manton:
You couldn't have picked French, huh?

Kerim Akgonul:
No, no, no, Turkish has the right solution for this moment. And Don is almost there. It smells great. I don't know if you guys can smell it but it smells awesome up here. All right. And Kara, what do we have?

Kara Manton:
Well, I typed Turkish. I don't know if that looks like Turkish.

Kerim Akgonul:
We have a winner.

Kara Manton:
What do you think? Oh, good.

Kerim Akgonul:
Kara, congratulations.

Kara Manton:
Yes.

Kerim Akgonul:
Well done. Excellent work.

Kara Manton:
Thank you.

Kerim Akgonul:
And that is actually pretty accurate.

Kara Manton:
Good.

Kerim Akgonul:
So can we actually hear what Kara just won?

Speaker 4:
Congratulations. Kara, you've won an all-expenses paid trip to the Innovation Hub.

Kara Manton:
Oh my God, amazing.

Kerim Akgonul:
Enjoy that, Kara. Thank you so much.

Kara Manton:
Thank you so much.

Kerim Akgonul:
Thank you so much for participating. That was great.

Kara Manton:
We'll just leave Don up here I guess.

Kerim Akgonul:
Yeah, he's fine, he's ... I want that omelet. I know that was fast, and I'm going to go through exactly what just happened here step-by-step. But what we just witnessed here is a totally unqualified person trying to make a Western omelet.

Don Schuerman:
All right. You know what? You know, Kerim, I'm going to take my pan I'm going to go home, you can finish the rest.

Kerim Akgonul:
You save that for me.

Don Schuerman:
Yeah, I will.

Kerim Akgonul:
Thank you, Don.

Don Schuerman:
Thank you.

Kerim Akgonul:
He was so close by the way, I mean, it's almost ready. In addition to that, we also saw an incredibly fast app build with Pega Infinity '23 leveraging the GenAI capabilities. And as you know here at Pega, we've been doing ... We've been building applications for quite some time so we have a pretty good idea of what are some core constructs that you need to pull together in order to get an application up and running. You have to build things around the case life cycle to find the process, you have to do things like identify the data model and the users. That's basically pretty well understood and it's well-defined within the application.

But traditionally, to do this, what we usually do is we grab a bunch of really smart people, we throw them into a room, and ask them to figure it out. We ask them to write requirements documents, we ask them to meet and discuss ideas. Often they write very, very long documents. And sometimes this can be wildly successful. Sometimes you get different results. And it becomes a problem when days and weeks turns into months and millions of dollars spent just to get a viable answer on how to automate a process. But with GenAI, all of that changes, all of that is gone. GenAI helps developers, helps this ... The business analysts to gather all the requirements and pretty much configure the phase one of the application. Now these smart people who understand the business can actually spend their time focusing on reviewing the implementation, and making configuration changes that are specific to the organizations, and customizing it to get it to work for the organization and to get it moving really, really fast.

Let's take a look at it step by step and see what we just saw. The first thing Kara did was to define the case type. And she leveraged the GenAI that's built into Infinity to actually gets the definition of the first phase of milestones and steps of the core process. Got it from GenAI, and already have it plugged directly into the product ready to execute. It just works. As we know, every single process is going to have some people, some personas that participated with it. So again, we leverage the GenAI capabilities to identify the different personas that are going to participate at different stages of the life cycle of this process and got that into the system.

So we've got the case life cycle definition, the personas defined. You all know that you have to have a robust data model that's defined for the application, both for the case definitions as well as for the secondary objects that's going to participate. Again, we leveraged the GenAI capabilities to capture a robust set of the data model for these objects. Now is it 100% correct? No, absolutely not. Is it really close? Yeah. And that's when those smart people come in and actually change ... Define the differences. You have a robust running application with the case, and the personas and the data model defined.

Now these objects that we define the data model for, often they're going to need to integrate with other applications within the enterprise. So we're going to have to build the integrations, we have to get the data model from that other system and map it into the data model in Pega. We just use generative AI to do all the mapping for us regardless of how complex the objects are. It just works. Now, you've done the integration, you got the data model, and you can start to run through some of the application, and you need to create test cases. Did I click? I'll click. And with GenAI, you can just actually automatically create sample records for all of your objects directly in the system within seconds. It will be accurate, realistic sample data, and as much as you want, you can have five, 50, 100, 1000, it doesn't matter, it'll just populate your systems for you. You can actually experience what the application is going to feel like and look like with realistic data. Now, you define the sample data and you want to see what it looks like for the end users. We'll leverage the GenAI capabilities to actually generate the user experience for those personas against that data model for those case definitions and its journey directly through the App Studio experience, and it just works. That's what we just saw when we looked at the screen at first. It wasn't English ... It was in English at first, but she basically knew that in order to win that prize, in order to win the contest, Kara had to translate that application to Turkish. Now, I'm still pretty good in Turkish, Kara isn't, but GenAI ... She's working on it. But GenAI actually leveraged ... Gave us the ability to just translate the application into any language that we want. We can translate it into Dutch, we can translate it into Hindi, we can translate it to French, Spanish, Japanese, Turkish, and it gets done in seconds. Now again, is it 100% perfect? No, but it's really, really close. And it can collect all the elements that are experienced by the users and translate them all for you.

Now, when you have an application up and running, right, when you're going through the processes, when you have the cases, and the processes, and the interactions going you're going to want to get some insights into what the operations look like so you're going to want to run some reports. And I know everybody has lots of BI expertise, and lots of people know how to write SQL statements but it takes a long time. With the GenAI that's built into Pega Infinity, you can just ask it a question. You can just say, "Show me these records, show me the loans, organize them by state, or organize them by loan type." The GenAI will automatically figure out what the question is, what the objects to go after, help us identify the SQL statement, will generate the insights report definition in Pega, will automatically determine the best way to represent that data, and show it to you as quickly as what you just saw. And you're going to see all of this. And all of these capabilities work because the GenAI is actually designed directly into the architecture of Pega Infinity. It's actually like baked into that layer cake that Alan was talking about as a core part of the product.

Now, during these presentations, I usually hesitate to show architecture diagrams but let's go for it, right? I think you guys will appreciate this. Now, you all know that Pega products is essentially ... The platform is a low-code app development platform for AI-powered decisioning and workflow automation. That's what we talk about as what the product does. And the way it works is through a studio like the App Studio, you capture the definition of the process, definition of the AI decisioning, this decisioning of all the automation components in constructs that we call rules. And then what we do is in the background we have a set of microservices that actually execute those definitions and turn those definitions into runtime. That's how the Pega works. There's rule times that define the purpose and then the services execute that purpose for you.

With the '23 architecture, we extended this set of capabilities to include a net new core component of the architecture called Pega GenAI, and it has three components to it. And for those of you who've known Pega for a long time will not be ... This will not be a surprise. We have a rule-type definition that allows us to capture the purpose of the prompt that we're going to engineer to leverage generative AI so we can capture the thinking. What is it that we're going to ask? The second component is a gateway service which is really like a traffic cop that figures out what service to use to execute that prompt. There are already tremendous numbers of GenAI services and large language models out there, and every organization will make its own decisions around which ones that they're going to leverage, including the last one basically indicates here. Including language models that you build for your organizations that we can plug in to get the work done.

And the third component, which is a critical component as well, is intended for being able to leverage GenAI with confidential data. It's a set of local models that help you process confidential data through GenAI without leaving your secure environment. And these are essentially net new components that you will see when you get a chance to see all the demonstrations. I just gave you really example of six, seven, eight of them, there's another 20, 30 of them out there. This is just the beginning, there's going to be many more. You will certainly, certainly see GenAI-based capabilities from many other vendors out there as well. You'll see things like they click on a button and it creates a form with two fields. You'll see them click a button and you get a chunk of code back like Alan was talking about. And some of these can be very useful for some simple projects. And I've already seen lots of the capabilities that look really, really cute. Really cute demos that target simple problems. And everybody likes cute, that's Skittles. We love cute.

But I have some bad news for you, and I know this is going to be hard for some of you to hear. Some of your problems, they're not that cute. They're kind of big. I've seen some of them they're scary. I've seen some of your processes and they need to be tamed. And it's usually not just one, they're a lot of these processes roaming around looking to cause trouble, and you know you need to fix them and optimize them. Well, that's absolutely normal. You run a major enterprise. You inherited an enterprise architecture that probably looks something like this. Your problems require a different level of expertise, a deeper sense of sophistication to operate.

And Pega, that's exactly what we do. We like to go beyond the bases, we like to understand your business objectives. We like to understand your processes, and we like to help you optimize them across this architecture. And to help us do that, to help us bring more value to you and to your efforts around driving optimization, we are thrilled to announce a brand new product called Pega Process Mining. Pega Process Mining is a brand new solution that we just announced. It essentially collects execution data of your processes across all of those applications that we saw on the enterprise diagram so that you can actually understand how a process truly runs. Not how you think it runs but how it truly actually goes through ... Across all those applications, across all those systems, across all those database records.

Process Mining provides analytics capabilities to help you understand the elapsed times, to help you understand bottlenecks, to help you understand where does a process actually get stuck. Where does it go down exceptional paths that are costly, and allows us to do analysis so we can define where to focus on in order to drive optimization directly to the organization? And as GenAI gets plugged into Process Mining, you'll be able to just chat with your process. You can ask, "What's wrong with our loan process?" You can say, "Where does it get stuck?" You can say, "Why does it go down these bottlenecks?" And it'll help you drive that analysis so that we can drive into operational efficiency outcomes and great experiences for our end customers. And you'll see this afternoon in the Innovation Hub.

You'll also see how we leverage this GenAI that's baked into the architecture, leveraged in other applications. For example, you'll see how we use it in our sales automation application which is really a top-ranked sales automation solution by Gartner. We use GenAI in there to make salespeople more effective by capturing action items and follow on assignments from engagements that they have with their prospects, helping making salespeople more effective.

You'll see us leverage GenAI in our customer service solution in many use cases. One of my favorite one is all around saving time for agents after they finish a call with a customer and automating the wrap-up summary steps, which usually takes minutes for agents to complete after every call before they can take the next one. Instead now it just basically shows them the wrap-up notes. The human is in the loop they say, "Okay," and it directly basically Registers that with the application saving minutes and minutes for every agent on every call. You will also see the value of generative AI in our industry applications.

For example, in our financial services crimes investigations solution, Pega GenAI makes a huge impact. Actually, we use the analytics AI capabilities to apply real-time event processing to alert for multiple systems around payments, and wire transfers, and deposits to create a visual to uncover suspicious activity. Then the financial service organization have to report them to appropriate government regulators complying with a varying set of formats that are specific to every single agency. If these suspicious activities are not reported in time there are some very, very high fines. And the application, the gene AI capabilities, helps our customers get those reports in saving tremendous money. And, of course, our heritage around some of these AI capabilities comes from one-to-one customer engagement.

And tomorrow, right here on this stage, you can't miss Rob Walker explain the AI revolution in different aspects of AI. And I'm hoping you'll all be here to hear from now ... From Rob, and I know it'll be a great, great presentation. You're going to hear from Rob. You're going to go to the Innovation Hub, and you're going to see all the demos, and you're going to ask yourself, how do I get my hands on this stuff? I believe the answer is Pega Cloud. Pega Cloud will make a big difference for your organization. Pega Cloud can help you save money, it'll help you save tremendous amount of time, and it will really reduce the operational stress. And Pega Cloud will help you keep up with the pace of change.

We all know that things are moving really, really fast. And this tectonic shift with generative AI is really our collective opportunity to actually reach for all those dreams that we always talk about. We cannot just be casual observers, we all need to get on board as active participants with all these AI capabilities. Together we can build for change, and the times certainly are AI changing. Thank you all.


Tags

Topic: AI and Decisioning Topic: PegaWorld

Related Resources

AI in Innovation

Unleash the power of AI innovation

Learn about enterprise AI and how it'll drive the future of business outcomes.

Why Pega

Why Pega?

Uniquely powerful software isn’t the only thing that sets us apart.

Share this page Share via x Share via LinkedIn Copying...