Launchpad at PegaWorld 2026

8:23
Well, thanks.

8:24
Well, good afternoon everyone and welcome to PEGA World 2026 and this our Launchpad session.

8:30
I'm John Hewen.

8:31
I head up Launchpad for PEGA and we have an action-packed agenda for you this afternoon.

8:38
In the next 90 minutes, we'll talk about big, exciting changes in our industry.

8:42
We'll announce and introduce some new capabilities that you've never seen from us before.

8:47
And you're going to hear from some great speakers and customers who are doing amazing things on the Launchpad platform.

8:53
Some that are developing powerful products that generate meaningful outcomes, some that are harnessing AI in regulated industries, some with agents that do work for their customers, and some who are bringing together a new set of tools to pioneer on the leading edge of.

9:09
And of course, we're going to celebrate some great achievements and hand out our second annual Launchpad Awards.

9:16
The competition has been incredibly fierce this year, and our judges wanted me to tell you that they've really appreciated all of your lobbying efforts.

9:25
I'm going to start by grounding our conversation with an overview of why companies love building on Launchpad.

9:30
Then I'll talk about the importance of outcomes in the software industry, and then I'll talk about AI and development and runtime as massive efficiency drivers.

9:40
Think about the solopreneur with a billion dollar valuated business, something that was a dream just a short while ago but now starting to come to reality.

9:50
And then think about how the tool that enables their success can help you and your business and maybe get you an office that's as slick looking as that with such a spectacular view.

10:00
Launchpad is a platform on which you can build and manage really any workflow centric B to B application.

10:07
You can make it AI powered.

10:09
You can make it consumable by an unlimited number of customers or subscribers.

10:13
And our mission is to make that as easy and cost effective as possible for you to build your app.

10:20
Launchpad gives you AI powered and low code development tools like blueprint, workflow automation, ready built UX design and integration capabilities that enable you to quickly and easily make something production ready.

10:33
And we're going to blow the doors off that a little more a little later on.

10:36
We're going to show you a demo with some AI powered development, some vibe coding, if you will, inside of Launchpad that I think is me stealing my own Thunder.

10:46
I'm going to let you see what that looks like a little later on to deploy your app, Launchpad gives you an enterprise grade, secure, scalable, managed cloud, your database, multi tenancy and more.

10:57
It means your app doesn't just run on a laptop, it's not local host and you don't need a big cloud OPS team.

11:05
When you click deploy, we take your app live on production ready, enterprise grade architecture.

11:11
Go ahead, add your customers, add your subscribers, and when you do, Launchpad gives you subscriber management, configuration, identity management, dashboarding, and lots of other stuff that makes it easy for you to onboard and administer your subscribers.

11:27
With Launchpad, your business can shift focus from DevOps and coding or even interpreting massive amounts of AI generated code to IP development and differentiation.

11:38
Go to market and the delivery of outcomes as you'll hear from our customers on our panel a little later.

11:45
Whether you're looking to build and deploy new enterprise grade outcome based solutions or whether you want to enhance pre-existing underperforming SAS solutions to make them into outcome generating solutions.

11:59
Or whether you need to replace tech debt laden, expensive labor intensive applications that you built from scratch or on another platform.

12:07
Maybe before you knew Launchpad, The Launchpad is the outcomes as a service platform to enable your success.

12:17
Now maybe you're thinking to yourself, why all this talk about outcomes?

12:22
Well, that's the big change that's so exciting.

12:24
In our industry.

12:26
For 20 years, companies bought software to help people do work.

12:31
Today, for the first time, software is beginning to do the work itself, and that is a very big change.

12:39
And when the work becomes the product, outcomes become the business model outcomes.

12:46
That's what everybody wants.

12:48
Not software, not tokens, not seats or users, not another login outcomes.

12:56
But maybe you're saying, John, sure.

12:57
But what I really want is I want AI.

13:00
Of course you do, but not for the sake of having AI.

13:04
Nobody wants more models or to spend more tokens.

13:07
You want AI that delivers an outcome.

13:10
Build my app, deploy it, run QA, find bugs, fix them, launch faster with AI and design time or vibe coding.

13:18
The outcome is your app, your workflow, your product.

13:23
But then maybe you're saying to yourself, yeah, that's great, but what I really want is agents in my runtime.

13:28
Well, of course you do, but not just because agents are the cool new thing or they look so darn cute.

13:36
Do you want agents because they perform work?

13:39
Agents that process claims, perform audits, generate leads, close deals, resolve support cases.

13:45
Agents that move work through an organization.

13:48
Agents that create value.

13:50
Agents that deliver outcomes.

13:55
Coming up next, you and me delivering outcomes.

13:59
As everyone in this room already knows, the world of software is changing dramatically right now.

14:04
It's already an overused phrase to say SAS is dead.

14:07
I don't know that SAS is dead.

14:09
Certainly that sentiment feels very May 2026 to me right now.

14:13
But I'm literally seeing large LLM companies post jobs to administer their SAS solutions, even including CRM.

14:21
But the story isn't about whether we're hearing the death toll for SAS.

14:26
It's about the fact that AI has fired the starting gun on something bigger.

14:31
Outcomes as a service.

14:34
If you're in consulting or a systems integrator, AI is compressing strategy planning and development cycles, skinnying implementation processes and projects, reducing the need for people, which is a pain for your traditional business model.

14:50
But it's also driving demand from customers to deliver not just the strategy strategy, not just the implementation, but to deliver the actual outcome, which is a really big opportunity for you if you'll grab it.

15:06
If you're a software company, the story's the same.

15:09
The decades of selling software to be used by humans, with pricing by the seat or the user or the license are evaporating.

15:16
That's a change.

15:18
But when the software and agents do the work itself, you have a whole new business model.

15:25
Outcomes as a service SAS sold access to tools.

15:30
Outcomes as a service sells completed work.

15:35
And the promise of AI is that AI and agents can help work get completed autonomously end to end, Not by having a bunch of agents trying to manage other agents, creating inconsistency that introduces Rick risk to enterprises and regulated industries, but with predictable agents guided and controlled by workflows.

15:59
That's an incredibly exciting agentic proposition.

16:03
And in this room today, we're going to throw fuel on that fire.

16:06
We're going to show you some amazing agentically delivered outcomes, some real life agentically delivered outcomes.

16:13
But as we do that, we're also going to acknowledge something else that's pretty important, that not every valuable outcome needs to be fully autonomous.

16:21
A tremendous amount of work in the world is still human centric.

16:26
Real businesses rely on judgement, approvals, exception handling, governance, trust.

16:31
And that's OK, because outcomes are still beautiful when they're delivered through systems that help humans move, work faster, smarter, and more effectively through an organization.

16:44
And This is why one of the most important enterprise models that's emerging today is a gentic.

16:50
With humans in the loop, AI does the heavy lifting.

16:55
Humans provide oversight where it matters because the biggest barrier to enterprise AI adoption today is no longer capabilities, it's trust.

17:07
Trust and maybe we may be increasingly token spend.

17:11
So let's solve for both of those on the right hand side.

17:14
Human in the loop closes part of the trust gap.

17:16
Workflow and deterministic logic Beside that close the rest.

17:20
Use AI in design time and where reasoning creates value.

17:25
Use deterministic systems where repeatability matters.

17:30
That puts trust and builds trust and keeps token spend focused.

17:36
Where intelligence actually matters.

17:38
Because intelligence without guidance, without control, That's not transformation.

17:43
That's potentially a pretty wild science project.

17:47
And sometimes that gets weird and expensive.

17:53
Bet you wish you had a workflow.

17:56
For many organizations, especially in regulated industries, the right combination of AI, workflow, deterministic logic, and human oversight becomes the ideal first step on their agentic journey.

18:09
But none of this happens with an LLM alone.

18:12
Outcomes require orchestration.

18:15
Outcomes as a service requires enterprise grade architecture, security, compliance, scalability, multi tenancy workflows, integrations, governance, operational controls.

18:25
I spend a lot of time with a lot of AI first companies.

18:31
Many of them have vibe coded up an incredible application that just runs beautifully on a laptop, often with mountains of AI generated code underneath.

18:41
It would take a brilliant engineer weeks to decipher and understand.

18:46
And then the enterprise shows up looking for those things that they require to ingest software into their businesses.

18:52
Things like, I don't know, a SoC 2 certification.

18:55
And then all of a sudden the application that one person built requires a full team to operate.

19:02
And maybe it seems obvious to the group in this room, but if outcomes become the product, everything else has to change as well.

19:09
Development, architecture, operations, economics.

19:12
Launchpad was built to enable outcomes from day one, not as a bolt on AI layer, but as a platform for building and deploying and orchestrating and governing and scaling outcome driven software businesses.

19:27
Really that's the case of the 40 year heritage of Pega.

19:34
And today we're making outcomes easier to build than ever by combining AI assisted development, again, what others call vibe coding with our workflow automation capabilities and production ready enterprise grade architecture.

19:49
We're going to show you that.

19:50
And what that is doing is helping it, helping make it easier for you to move from idea to workflow to deployed agentic application to customer delivered outcomes faster and easier than ever before.

20:06
And now?

20:08
To show how all of this comes together, I'm going to invite Steve Bixby, our senior vice president of product engineering, to the stage.

20:14
And if Steve's willing, yeah, you can clap for Steve.

20:17
That was great.

20:20
If Steve's willing, I'm going to give him a bit of a mission.

20:23
I am going to ask Steve, Steve to build an agentic application using our absolute latest capabilities to solve a problem that has affected my family.

20:35
Steve, this is my sister Sarah.

20:38
You can probably see the family resemblance.

20:40
She's a cartoon and approximately probably 22 younger than I am.

20:43
But Sarah is also just a small town girl living in a lonely world somewhere in middle America.

20:49
Steve.

20:50
And last week her dream home came available in beautiful Bloomington IL or Indiana.

20:56
Sorry Bloomington, IN.

20:57
And she booked Steve a showing and called her bank to get a mortgage pre approval letter.

21:04
She quickly sent through all of her financial documentation.

21:07
But then it took two days Steve for the bank to run their processes and send Sarah her letter and well, Steve.

21:15
By the time Sarah's letter came, the house she loved had sold to someone else.

21:21
Bummer.

21:21
Imagine the disappointment.

21:23
Unbelievable.

21:24
So Steve, what I'm hoping is that maybe you can suspend your role as the senior Vice President of product engineering at Pega and instead take the time to build an app for banks with an agentic workflow that shortens the mortgage pre approval process from days to seconds.

21:40
Saving the banks, lots of manual processing and delivering an experience that looks like this.

21:46
An agentic app with a conversational interface that works nicely for people like my sister Sarah.

21:52
Steve, are you up for it?

21:54
Yeah, That sounds like a lot, to be honest.

21:57
But yeah, of course, that's why we're here.

21:58
Right.

21:59
Let's do some live software.

22:02
So, yeah, I did happen to just, you know, found a new company in the last week.

22:08
Wow.

22:08
It does exactly what we're looking for.

22:10
I call it Lone Ranger.

22:11
That's amazing.

22:12
And yeah, we asked ourselves, what if we built an agentic application that accelerates processing and delivers an amazing customer experience for mortgage pre approval?

22:19
This is an unbelievable coincidence.

22:21
It's like, right, right.

22:22
Exactly what your sister's looking for.

22:23
So if we can just get some subscribers in Indiana, we're good to go.

22:28
We'll work on it.

22:29
Let's do it.

22:29
OK.

22:30
All right.

22:30
So sweet.

22:32
We're on the screen.

22:34
No surprise.

22:35
If we were building The Lone Ranger app, we would start in blueprint, right.

22:40
Right on launchpad dot IO, which is exactly what I did, gave it a name, gave it a description, chose the company and it outputted, you know, a series of workflows.

22:52
We're going to focus on the pre approval workflow, obviously.

22:55
Thank you.

22:56
So let's drill into this a little bit.

23:00
We're going to see a lot of the first three steps as I go through this because we're going to have to keep testing some of the changes we're making.

23:05
So we're the first step is loan details, then property details, applicant details.

23:10
There's an eligibility review stage that we're going to try to use and then there's a alternate stage for denial.

23:17
I'm not going to go too much into the rest of the blueprint other than to just lay, let's preview this thing right?

23:22
This is one of the most powerful things about using Launchpad IO and the blueprint capability, and that is being able to have some certainty of the outcome that you're actually going to be delivering because you can see it right away, right?

23:37
So right away I'm now seeing this is what an approval workflow would look like to someone using The Lone Ranger application at a bank in the back office.

23:45
This is like a power user that's going to be sitting and doing these things all day.

23:49
And you can already see that it's got the pre approval case open.

23:51
There's the loan details, there's the property details, and there's the applicant details.

23:57
I click submit, it resolves it.

23:59
I can do this with all the other case types.

24:01
And hopefully, as you know, you can see what is this going to look like to someone on a mobile device, right?

24:08
And you can actually start to play around with, make changes to experience that.

24:12
What if I wanted to see what would this look like to a consumer on a website using this?

24:21
All right, now this is showing the provider's website as opposed to a subscriber's website, but nonetheless, this is that pre approval flow right here on a website.

24:31
All right, this is using the Launchpad web component.

24:33
We're going to see some more of that in a few minutes as well.

24:36
Finally, you can always preview the agentic experience.

24:40
This is using the Launchpad provided agentic case processing agent right?

24:47
So that you could actually deploy something like this right on your website and say hey I need a pre approval approval.

24:57
I have a really hard time typing pre approval.

24:59
It's proved to be a weakness I can't overcome.

25:03
So it's now it's actually running the workflow identifying, OK, this is the pre approval is the case type that we want.

25:10
It's asking for some data.

25:11
So I'm just going to say how about 500 K for 30 years for and I'm going to put a lot down, I'm going to put 200K down.

25:20
And what this is doing is actually using our agent to take that information, create a case, post information to the case, execute the workflow to inform the next prompt, right?

25:35
And at any point, you can always peek behind the scenes and see what is this actually doing, right?

25:40
And I encourage every single one of you to try this on your own if you haven't already.

25:44
Just go to launchpadioorpega.com/blueprint, create a blueprint, create, look at the conversational agent, have a chat with it, give it a call, like, try this, call the number, enter the code.

25:56
You can have a chat with it, right?

25:58
And I'm going to show you the next thing I want you to try as well, which is this new thing at the bottom here, which is MCP agent.

26:04
So say you don't want to use the Launchpad provided agent, but instead a subscriber already has their own agent.

26:12
And they're asking, hey, how do I plug in this SAS software to my existing agent?

26:18
And the answer is model context protocol, right?

26:22
So right from blueprint, every one of you can create a blueprint and then come in here and see how can I plug this into an existing agent.

26:30
I'm going to use Claude in this case.

26:31
So I'm just going to copy the URL I'm going to toggle over to my Claude desktop.

26:37
I'm going to hit customize connectors, add custom connector.

26:45
I'm going to call this my Lone Ranger blueprint and I'm just going to paste the URL right from blueprint and it should there we go.

26:55
Creates this thing for me.

26:56
Oops, I spelled loan wrong.

26:59
We're going to live with it.

27:01
All right, so second company lay on Ranger lay on.

27:07
OK, so I'm just going to say that type the exact same thing I typed in the blueprint preview, right, And I'm going to say I need a pre approval.

27:17
What this should do is identify that it has this plugin now, right?

27:22
And hold on tight.

27:24
Create approvals, searched available tools.

27:28
Did it find it?

27:32
Looks like it.

27:32
So hold on, let me do 500 K, 30 year, 200 K, and if it doesn't give me the case ID, I'm going to ask it for the case ID just so we know that this thing is actually using that same blueprint.

27:49
Yeah, there we go.

27:50
It's using the the plugin.

27:53
So this is amazing, Steve, like, you know, I'm not using a Pega interface or a Pega agent.

27:59
I'm now running this Pega workflow, this Pega agent in another system, in another agentic chat, your agent of choice.

28:07
I've done nothing except visit Launchpad IO and create a blueprint, right?

28:11
And I'm like, I can do all this crazy stuff, which I think is important to show all of the different ways that people want to interact with the software that we're building.

28:22
Like I can very confidently say SAS is not dead.

28:25
SAS is absolutely required so that you have this capability on the back end for your agents and for, you know, any of the things that people are vibe coding on the front end, etcetera, etcetera.

28:36
So let's go back.

28:38
One last thing I want to show in oops, sorry.

28:42
One last thing I want to show in Blueprint is something new that if you've not seen, definitely check it out.

28:51
John talked about all the things that Launchpad gets you.

28:54
He's going to kill our astronaut.

28:58
Suddenly it occurred to me.

28:59
He's over here screaming.

29:00
Does everyone feel less stressed now?

29:03
I do, right.

29:06
Thank you.

29:09
Well, as I was saying, yeah.

29:13
New feature.

29:14
I'm just kidding.

29:14
Thank you.

29:17
As I was saying, there's a new feature on the last step of Blueprint that's going to show you all the capability that makes Launchpad so much more powerful.

29:25
All the things that John was just describing about, you know, making sure you've got the security and encryption capabilities, that you've got regulatory compliance, that you've got operational controls in place.

29:37
All of that is here as part of the Launchpad platform and right available available for you to look at right here.

29:43
OK, the next step, let's just build this thing.

29:47
Truth be told, I already did import this blueprint, but whoops, that's not what I want.

29:55
But if you haven't brought a blueprint in the Launchpad recently, it's massively improved.

29:59
It literally takes a few seconds.

30:01
It's just like bam bam, which is pretty awesome.

30:04
The way you do it is you just click new application, choose your blueprint from the list, and you're off and running.

30:09
That's exactly what we did here.

30:10
So now I have this Lone Ranger application, something different about Launchpad Studio here.

30:17
We've got a very prominent AI assistant.

30:19
We were wondering if it was going to say good afternoon, John, look at that.

30:23
The AI even knows that we're in Las Vegas as it is now.

30:26
Afternoon.

30:26
Yeah, this morning we noticed that it said good morning.

30:28
We're like, does it always say that or is that anyway, I'm going to use this to refine this application and I'm not even going to, or I'm going to try not to use any of the navigation on the left.

30:42
I'm not going to look at any rules.

30:43
I'm not going to go to the case designer.

30:45
Like let's just see if we can use the AI assistant to enhance the workflow.

30:49
We're going to run a couple scenarios, but before we do that, let's preview the application so we can see it running for real.

30:56
It should look just like our blueprint, maybe TikTok.

31:06
All right, when doing agentic demos, not that this part is agentic, but you have to find time, ways to kill time as things are running in the background.

31:14
So you're going to see me doing a lot of dancing in a few minutes.

31:18
All right, So if we create a new pre approval, yeah, we get the information about the loan, we see the information about the property, and then we see the information about the applicant, right.

31:29
And actually, I'm going to go back, let me do another one because I want to have the data populated for when we come back through here.

31:34
Sorry, so I'm going to fill the form with AI TikTok TikTok next fill the form with AI next.

31:47
And right now we haven't done anything.

31:49
I'm just setting this up so we can come back to this case in a in a minute.

31:53
All right, so we've got the pre approval.

31:55
Looks like it's gone all the way through.

31:57
It said, Yep, you're pre approved.

32:01
You can see the requested loan amount here over here.

32:03
We call this the summary panel on the left.

32:05
You can see the down payment, but I'm thinking it would be nice.

32:09
Maybe I should have done this in blueprint, but it would be nice if we could see the down payment percentage on the left like that seems like a summary field that someone would definitely want to know about with a loan pre approval.

32:21
So let's take that fairly simple UI change a little bit of a calculation that has to take place.

32:25
So let's try that.

32:28
All right, so I'm going to drum roll add the calculated down payment percentage to the summary panel.

32:39
All right, here we go.

32:41
Now I've run this enough times to know that it is a little bit non deterministic and there are times where it gets off track, but usually I can recognize when it is.

32:51
Yeah, I mean the great part is while it's while it's building, if you see anything on there that's going a little bit scooters, you can always grab that and redirect.

33:00
Of course it's having something that goes off and just generates a whole application for you and then you get scooters.

33:05
Awesome.

33:05
So it found yeah, it found a skill.

33:08
It's off and running.

33:09
I'm I'm fairly confident this is going to do what it's supposed to do.

33:12
So while this is going iterating, trying to figure this all out.

33:17
Good news, John, we got our first subscriber.

33:20
Oh my gosh, I knew there was demand for this.

33:22
Just got a notification, so I'm going to bop over to the subscriber portal and and you can see I've already got my test subscriber in here, but I'm going to add the new subscriber.

33:33
Good news also there in Indiana.

33:36
So that could help your sister, right?

33:38
Indiana Loans is what we're calling it.

33:42
That's a good one.

33:42
And yeah, we'll give them an ID.

33:45
We're going to say their representative or contact is Henry [email protected], just to show how easy it is to create a subscriber.

33:58
And so now we've got Indiana alone set up and this is sort of ongoing as it's provisioning the resources.

34:07
It takes a few seconds and then that will be done.

34:09
So that was a stall just to show you that we have subscriber capability and we're going to push changes here.

34:15
Let's go back and see if it's done.

34:17
It is still going OK now as it's taking its time, Steve, like we're seeing it's doing all of this work.

34:23
Just give us a sense for if it was a person who was doing this work, how long would that take versus how long this is turning around?

34:30
Yeah.

34:30
So that's a great question.

34:31
And in this case, I think a person probably could have pulled this off in about the same amount of time, right?

34:36
This is a fairly simple change.

34:38
It created a decision rule.

34:39
It did a simple calculation to determine the percentage.

34:42
It created a calculated field and then put it on the view, which is the summary panel, like.

34:48
So if you know what you're doing, this is piece of cake, I think with my LSA training.

34:52
Yeah, if you know what you're doing exactly.

34:54
But if you don't, this is brilliant, right?

34:56
This is amazing.

34:57
And actually, we've started to see that even the people that know what they're doing, like my team that knows this stuff cold, is starting to use this because it's just easier, right?

35:05
And you can do other stuff while it's happening.

35:08
I'm sure any of you that have been using AI to write code, which we do a lot of at Pega, you have a lot of time to do other things while things are running, right?

35:18
And so that's another benefit that you get.

35:21
We're going to do some slightly more sophisticated scenarios in a second, but let's test this one.

35:25
Let's just launch the preview again and just see if that percentage calculation actually is showing on the summary panel on the left.

35:36
Here we go.

35:37
So this is running the math.

35:38
We're just going back.

35:39
This is pulling in the percent down, putting that on the left hand panel.

35:44
Yes, exactly, Not before, exactly.

35:48
Another plug here.

35:49
If you haven't seen Blueprint imports now create landing pages for every one of your case types with reports that you can define in Blueprint, right.

36:00
So that's another really great new capability in Blueprint as well.

36:04
And this is what it looks like.

36:05
I'm going to open this last one.

36:10
That was that we just did and there we go.

36:12
I did this 4 minutes ago, but now I also see the down payment percentage on the left.

36:16
So this thing worked.

36:17
That's great, very exciting.

36:20
Thank you.

36:21
Thank you.

36:21
I'll flap for it.

36:22
OK, All right, so let's keep going.

36:25
Let's go back.

36:26
Let's use that same field now to actually determine eligibility.

36:31
Like we need some kind of eligibility check here.

36:33
We're not just going to let anybody get pre approved.

36:36
So I'm going to add this simple rule.

36:38
Let's turn this on now.

36:42
If this down payment percentage is under 10%, let's make sure the pre approval goes to a manual human approval.

36:49
All right, so this is going to hopefully go ahead and add a business rule, inject it into the workflow, maybe tell it to switch stages.

37:07
There are a couple different ways this could go Teddy's telling me this is going to work That's great thank you All right while this is going let's go take a look at what's Indiana Loans has been up to now that they're a subscriber and they have the first version of this application.

37:23
So your, your sister wanted to see an experience right from her mobile while she's looking at a house where she can have an agentic chat experience.

37:35
So we built this, we vibe coded this Indiana Loans website and but it has launchpad on the back end, right?

37:45
So that we have deterministic experience and all the security we need, all the compliance considerations.

37:52
And right there I've got this agent now this is a, an agent that Indiana Loans built, right?

37:57
This is not the Launchpad agent.

37:59
This is their custom chatbot on their website.

38:01
So this is like what we saw with Claude prior.

38:04
Exactly.

38:05
And so we're going to say same thing we've been saying all along.

38:08
I need a pre approval.

38:10
Look at how hard that Sarah is working to get this pre approval.

38:15
All right.

38:16
So hopefully this is going to hit that same workflow and ask for those same initial set of fields about the loan.

38:27
And then I'm going to tell it, Oh boy, come on.

38:31
Woo, woo Hoo.

38:33
OK, so loan amount, down payment term, same thing.

38:36
So I'm going to do the 500 K 30 year.

38:40
And I think we know what's going to happen now because we've seen this three Times Now.

38:45
We saw it twice in Blueprint and now we're seeing it for real on a subscriber's website using the capability from The Lone Ranger provider.

38:56
And there we go.

38:56
And now it's on to the next step.

38:57
And I can continue to provide this information and walk through the workflow.

39:02
Cool.

39:02
OK, let's go back, see how we're doing.

39:07
Why can't they go back?

39:08
There we go.

39:11
It looks like it did it.

39:14
It did decision rule.

39:15
It created an approval notes field, updated approval view, modified process.

39:21
So it looks like it worked.

39:24
Let's check it out.

39:26
Preview.

39:28
And actually while that preview is loading, I want to go back and just let's release some of these changes.

39:35
Let's see view the contents.

39:37
We're going to commit all this stuff, OK, and then we'll go release this thing and OK, so back to the preview.

39:57
Where's the preview?

39:58
So what you just did there you?

40:01
Took the work that you had just done merge that into a branch that's now back into the main branch and then you've released that and made that available.

40:06
Yeah.

40:07
So I want, because I want to show it in the Indiana Loans website, right.

40:09
So here we are in the the provider preview app.

40:13
Let's do our pre approval.

40:15
I'm trying to remember what we're doing here.

40:16
We're doing a loan amount that's less than that has a down payment that's less than 10% so that we have to go to an eligibility review.

40:25
That's the test we have to do.

40:26
OK.

40:27
So yeah, I'm going to do 500K, but I'm only going to put 20K down.

40:35
I'm going to click submit.

40:36
I'm going to click submit again, and let's see.

40:43
Awesome.

40:43
OK, so it kicked us to an eligibility review stage and now it's presenting this information.

40:49
There's the approval notes that we saw that it created, and I'm just going to say no, I'm going to reject this.

40:55
And now we have that option.

40:57
And if I go through again really quickly and do 500K with 200K down 200,000, Steve's going to be at the high rollers table tonight big time.

41:09
That's actually where I got that down.

41:13
And this should Yeah, good it goes.

41:14
So it goes all the way through.

41:15
Awesome.

41:16
So the first two vibes changes that we made or AI assisted change we made worked.

41:20
But I want to do something a little more interesting now, something that you wouldn't just sit down, make the changes yourself, save, test and and go because it's just a little more sophisticated.

41:31
So this one, wouldn't it be cool if, because we're going to need your W2 anyway in this loan process, if we added a field to upload AW 2 and then used AI to extract the information that's necessary for the rest of the pre approval right from the W2 form, right?

41:50
Seems very logical.

41:51
Feels like feels like a good outcome, right?

41:54
How do you do that?

41:55
Like I don't know, I think some people would do it one way, others might do it a different way.

41:59
Let's just ask AI to do it for us.

42:03
Now a little bit of AI hygiene here.

42:05
I'm not going to use the same conversation.

42:07
I could, but this is a different topic, right?

42:11
And I've, this is something that we talk about all the time internally.

42:13
It's a different topic.

42:14
Don't don't just clutter your context window with all this other stuff.

42:17
So just a little hygiene.

42:18
I'm just going to create a new thing and I'm going to say, let's try to extract some data from the W2 document rather than making the user enter it manually.

42:31
Let's add the W2 field on the first step of pre approval.

42:35
Then use Gen.

42:35
AI to extract the applicant name, employer name, and annual salary.

42:42
Jenny, I always, 100% of the time, sometimes it just says Jenny, That's a lot of stuff.

42:50
That's a lot of stuff in there, Steve.

42:52
We just asked it to do a lot of stuff.

42:53
But that's, that's what we want.

42:55
That's the outcome we want.

42:56
It seems like a very obvious outcome that seemed just a couple years ago, like something that was so difficult.

43:03
Oh, I need a OCRI, need to do some integration with this.

43:06
Like it seems like a big deal.

43:08
I don't think this is a big deal.

43:10
We'll see if this works.

43:13
OK, So I'm going to make sure that this thing is sort of on the right track here.

43:18
Let's see.

43:20
So what we've seen Steve, just like I recap this, some of this stuff like you've done the AI and design time to build the capability to see what the pre approval amount is.

43:30
And now we've created another capability that's actually using the agent to determine whether the pre approval, like the down payment is enough to warrant an automatic pre approval, you know, to warrant a pre approval, you know, automatically or if it's not automatically pre approved to escalate that to a human being.

43:52
Yep.

43:53
And now we and we've got this thing cooking and now we're pulling in the W2 so that we can actually just pull the data directly from the W2 to be able to drive that process.

44:03
Cool.

44:04
So while that's going, let's go take a look at what's going on on the Indiana Loans website.

44:09
We already saw the agentic chat, but I want to see what is that consumer Launchpad web component experience look like, like we saw in the preview in blueprint, right?

44:20
So we've, like I said this, we've, I coded this website, built this loan calculator, right, where I can change my down payment amount and I can see the percentage changing over there on the right.

44:30
Let's go less than 10%.

44:33
I wish I could change the interest rate.

44:35
That'd be pretty cool.

44:38
Zero.

44:40
So and now when I click claim this treasure, the Pega or Launchpad web component is pulling the information from that little widget and populating it into the workflow, right.

44:52
So it just pulled the loan amount, the down payment as well as the term.

44:57
And now we're running the workflow using Pega generated UI themed to match the Indiana Loans website.

45:05
All right.

45:06
And so as we step through this, now, let's get to the end.

45:12
Expedition funded.

45:13
I did not expect that because that rule should have been in place.

45:17
I'm going to try it one more time.

45:22
Let's go 2% down.

45:25
I'll try it one more time.

45:27
Was it the math?

45:29
There's a little error there.

45:31
Ignore that pending.

45:35
OK, that's, that's what I expected.

45:38
So it says your expedition is staged and ready.

45:40
We're completing an eligibility review, which is that step that we put in, you know, watch your inbox for the green light.

45:48
So that's interesting.

45:52
If we then Jack up this down payment to get it over 10%, I expect that we'll see the the opposite and this thing should be approved, expedition funded.

46:04
Great.

46:04
OK, so so that's great.

46:07
Going back to agentic end to end automated processing by the agent when it's being driven by workflow governance, workflow governance and deterministic logic.

46:19
So it came through and said this looks good, I'm going to pre approve that.

46:22
Then we got into the prior example was the agentic with human in the loop, where again, it was looking at it from a deterministic logic perspective and saying OOP, that doesn't meet the criteria that we need to have and that passes on to the human being.

46:33
But the workflow and the AI and the agent is doing, you know what we talked about the heavy lifting and then pushing that over for judgement decision making, you know, the trust capability back over to the human being.

46:46
Very cool, right?

46:46
Thanks for stalling while I push those changes.

46:50
It looks like it did work.

46:54
Summary of changes Jenny, I powered W2 data extraction with five rule changes Jenny iconnect, Jason data transform a view and automation and let's preview this see how it works.

47:09
Here we go.

47:12
Last one, thank you for thank you for your patience.

47:15
A question as you go through that, Steve, like the I see that Indiana Loans has some pretty Hee Haw language, you know, in the agent and in the conversation that it's having like because this is built on Launchpad and it's coming through Lone Ranger, does every bank have to have Hee Haw language or can I be a real life bank with real life customers?

47:33
Yeah, good, good question.

47:35
I think that's one of the real benefits of having the MCP enabled work flows set.

47:41
You can build the agent of choice, you can have the personality you want and that's what Indiana Loans has done.

47:46
If we had another bank that uses the same Lone Ranger application but didn't want to have your had some other personality, they want to have some real serious tone, they would build an agent that could do that.

47:56
So you have lots of choices.

47:59
We had a lot of fun with the AI just, you know, making a Hee Haw like you said.

48:04
So here we are.

48:05
I'm previewing this.

48:06
The W2 field does show up.

48:08
That's awesome.

48:10
Drinks are ready.

48:13
I'm going to pull this sample W2, upload it and you can see this is a example for Susan Quintana from Albuquerque.

48:26
She works at Westway Books and made 16 five and now what I expect is on the last step where it asks for the applicant information, I expect that those fields will be populated.

48:39
So they're, I can already see them there.

48:42
They are awesome.

48:43
So Susan Westway Books and that's her annual salary.

48:46
So this thing worked.

48:48
So this was a much more sophisticated scenario.

48:50
It took about the same amount of time, looked like.

48:53
How long do you think that would have taken if I was a human being doing that particular exercise?

48:57
Again, if you, like, knew every step of the way and you'd planned it out, I think probably, you know, 10 minutes, 20 minutes if you knew exactly what you're doing.

49:07
Realistically, I think this is the kind of thing that people take days to do.

49:10
Right.

49:10
Right.

49:10
I mean, how many times have we asked, oh, just can you just only upload AW, 2?

49:13
And they're like, yeah, sure.

49:14
And then week goes by and you're like, you've done yet another cool thing about this is that you can ask it.

49:22
You don't have to have it building stuff for you.

49:23
You can use it as a way to make a plan, right?

49:27
So you can have a conversation with it in the same way that you would with any vibe coding tool.

49:31
Say, let's talk about this first before we go ahead and build it for the sake of time.

49:35
I'm just like Bill, Bill, Bill, because I want you guys to see a result.

49:39
But that's that is a great feature of this as well.

49:42
All right, what was I going to show?

49:44
All right, so let's go finally show the website.

49:46
And then there's one last feature I want to show before I I turn it back to you, John.

49:50
So let's go back to the base camp and we want to show the W2.

49:57
Thank you should be here.

50:01
Boom.

50:02
All right, W2 is here.

50:03
So now all those changes that we made have been pushed released and are now available to subscribers so that we now can see all of this, this exciting stuff happening here.

50:16
And you'll see in a second like it was really fun building this website and integrating this stuff in.

50:22
We want you guys to play with it.

50:23
We're going to put AQR code up in a second so you can just have have at it try to break the agent.

50:27
Do whatever you want.

50:28
Definitely give your feedback.

50:29
It's in your best interest.

50:30
Sure, the initial part of the presentation.

50:34
I think you have to keep good applicable to Launchpad as well as the features of the blueprint.

50:43
However, the last part of my going if I'm a Chrome that is specifically applicable to you're going to have to show up tomorrow and find out.

50:56
Yeah.

50:56
What is the vibe coding that Steve just showed the AI assisted development?

51:00
Is that only on Launchpad?

51:02
As of this second, you would only know about this on Launchpad yes, but yeah, you will see tomorrow like and you'll see on the innovation hub for lots of cool stuff happening with AI assistant in Peg Infinity.

51:17
You're not going to see launchpad stuff outside of this session, but we're going to be here all day all afternoon.

51:24
If you guys want to talk more about this yes yes, this is not a Peg Infinity session.

51:41
The whole rest of the week is OK Last thing I want to show you guys is Oh my God, is another cool feature that you can, you're actually going to see a lot of with with in Peg Infinity.

51:54
I think you'll see this in Krim's keynote tomorrow as well.

51:58
It is the ability to change the UI from the end user experience.

52:05
And just bear with me for a second.

52:06
I have to to work make this work.

52:08
I have to blow away the cookies.

52:12
Let's see.

52:18
So we call this the UI designer, and it's going to be available to you right from the Launchpad preview.

52:26
If the Launchpad preview loads, oh, come on, one more time.

52:35
It's not going to work anyway.

52:39
You can imagine this very white screen.

52:43
I have a video of it that I can show, but I, I think it's better to take my word for this.

52:47
Crim's going to show some of this tomorrow in his keynote with Peg Infinity.

52:51
That same capability is coming to Launchpad, right?

52:55
And you can do things like change the navigation of the portal.

52:59
You can change the theme right from here, Go to dark mode, go to any other theme, and it changes right in front of you in real time.

53:05
You can change the layout of the case forms as well.

53:08
And what you'll see tomorrow is the ability to actually basically do what live UI did for you in Pegas traditional UI where you can just swing your mouse over the screen and just start making changes.

53:20
So some really exciting stuff coming.

53:23
Thank you.

53:23
That was fun.

53:24
We we built it.

53:27
I thank you for your patience.

53:28
It is hard to like know this is actually on the right track when it's sort of non deterministic like that.

53:33
And we'll share with you guys that these things, the MCP capability and Steve will talk about it a little more in a moment.

53:39
The AI assisted development, the MCAP capability is real tech.

53:43
What we're showing there, it is still early stage.

53:45
So we are rolling this out as we walk away from Pega World to early adopters.

53:51
If you are interested in trying out and playing with and participating with us in the AI assistant at the MCP capabilities as as Steve said, we're in the room for the next three days.

54:02
There's an opportunity to talk to us about, hey, I want to be an early adopter and try that and play with you guys and then we'll make that available to all very shortly.

54:09
Cool.

54:09
So I think if we go back to the PowerPoint, yeah.

54:15
So as I mentioned, if you want to, if you want to play with this little website where you can do the the web embed or the agentic chat, go ahead and do that.

54:25
John's going to show it again in a second if you miss it, if we go to the next one.

54:30
So this is what we, what we showed essentially in this demo is, is, you know, the AI assisted development, which is critical and frankly, I think table stakes at this point in order to remain relevant in 2026, right?

54:45
I think this is where everything's going.

54:47
I think we have a potential leg up here because we're not generating code, right?

54:53
No matter what, when you end up with code, you have a hard coded set of logic that you then have to go revisit, rework and maintain.

55:03
When you have a model driven no code experience, you're capturing the intent and the outcome that you want.

55:09
And we are providing the code right.

55:12
And it's not going to be determined non deterministic.

55:15
It's not it's, it's going to be accessible, it's going to be secure.

55:18
It's going to have all the things that we promised that would be otherwise very difficult to maintain over time in a vibe coded experience.

55:26
Just like we all know we've many of us have had to maintain code based applications for years and years and years.

55:32
It's, it's not easy.

55:34
And while there's some awesome technology out there that we're using internally for sure to write code, I think this is a better approach.

55:43
The UI designer, again, you're going to have to, you know, take a peek at Crim's keynote.

55:47
And then there's a bunch of other stuff that we've been working on over the last six months for Launchpad as well.

55:53
Lot of these ideas and enhancements came from real use by providers.

55:57
So thank you for your feedback.

56:00
It's been great getting to, you know, continue to grow the business a little bit, get more and more providers with more subscribers so that we can really drive the road map based on your experience.

56:10
Well, thanks so much, Steve.

56:11
This has been amazing AI assisted development.

56:16
You created Steve, an agentic app that massively reduces processing time for banks and delivers a great experience for people like my sister Sarah.

56:24
Just imagine how happy Sarah was when that house came back on the market and she secured it funded through Indiana loans.

56:33
Truly a spectacular outcome.

56:36
So there's one last thing that I want to make sure we call out that if you're building for outcomes, there's something else that changes too, and that's your economics.

56:43
Because if work becomes the product, outcomes become the measure of value.

56:48
For Launchpad, the currency of outcomes as a service is Launchpad units.

56:53
Little bits of work reads and writes of data that are used in the production of outcomes.

56:59
If your outcome is processed loan pre approvals, we make it easy to see how many Launchpad units it cost to deliver loan pre approvals.

57:08
Know your costs.

57:10
Set your price for your outcomes.

57:12
Different amounts of work drive different levels of consumption, which enabled different economics.

57:18
More steps, stages, data passed around equals more launchpad units.

57:22
Less equals less.

57:24
Based on this, you might price outcomes differently across customers or use cases depended on the amount of work that's required to deliver their outcome.

57:35
And now your business model aligns directly with value creation.

57:41
You can price by outcomes.

57:43
You can scale by outcomes, You can optimize your margin by outcomes.

57:47
And that is a fundamentally different way of thinking about software.

57:52
So I'm going to start wrapping this up here by saying the next generation of technology companies won't simply sell software.

57:59
Outcomes will be their product.

58:01
The winners won't be companies who generate the most features, they'll be the companies that deliver outcomes the most effectively.

58:09
This is the change.

58:10
For years, software helped people do work.

58:14
Now software is beginning to do the work itself.

58:17
And that presents a whole new opportunity for everyone in this room because what people want is outcomes.

58:24
Outcomes are what Launchpad is built to deliver.

58:27
And today we're here to help you move from idea to workflow to deploy to Gentek application to customer delivered outcomes in ways that you could never achieve before.

58:40
The Launchpad team is going to be in this room for the next three days.

58:42
Talk to us about how this evolves your business.

58:45
And now I'm going to turn things over to Jason Macharelli, our Vice President of Go to Market and head of Pega Ventures.

58:52
And Jason's going to bring up our panel.

58:54
I'll show you can clap for Jason as he walks.

58:55
So why not?

58:59
And I am very much looking forward to the opportunity for you to hear from Brad from Fiello and Arun from Avansis and Adrian from Ficent and ILCO from Talent League and BPM company.

59:10
And I'll let Ilco explain all of that to you and to introduce themselves to you as they come to the stage.

59:17
We're going to watch this quick little video.

59:23
Ivansis is a technology company.

59:25
We provide digital transformation services and products to the financial services community and one of our initial products is Trac EI.

59:34
Trac EI is a case management application that helps banks manage and process exceptions and investigations related to cross-border payments.

59:44
Yep, all good.

59:46
I'm Adrian Murray, I'm the founder and CEO at IC Technologies.

59:50
We are an applied AI software company based in Toronto, Canada.

59:54
Our platform, Biz AI, enables regulated enterprises to turn unstructured content into data that drives automation.

1:00:03
So at Fisent, we partnered with Launchpad to develop a regulatory testing application called Reg ERI.

1:00:11
The application is built to model the testing process by which banks and other financial institutions will evaluate the effectiveness of their consumer disputes according to Regulation 8.

1:00:26
And so the Launchpad workflow drives the end to end disputes testing life cycle and fizz and Biz AI is embedded throughout the workflow to analyze the evidence that is provided as part of the disputes process.

1:00:41
My name is Elko Fischenka, I'm the CEO and founder of BPM company.

1:00:46
Last year, we started to work with Packer Launchpad.

1:00:49
First application that we built is Tenantly.

1:00:51
Tenantly focuses on the HR processes in your organization.

1:00:57
It helps managers and employees to build their own career, facilitate their career and guide them to their career.

1:01:04
This application really helps you guide you through the process, not forgetting any steps in the process, facilitates managers to run the process correctly and it helps employees to really think about the career, what to do, what to develop with themselves, how to work on the competences and how to set their personal goals.

1:01:30
Brad Bellman, COO of Yellow, our core offering Fellow Loyalty, is a very robust loyalty solution.

1:01:38
The solution has recently undergone implementations of AI where the platform will provide recommendations for you in terms of what program should be run, how are your current programs are running.

1:01:51
It allows for you to set up an incentive program using our AI agent.

1:01:55
Halo partnered with Launchpad last year to re platform our core loyalty and incentive product that we have.

1:02:04
Part of that implementation was the building out of a loyalty designer, which is an AI agent that has a discussion with you, ask questions, challenges you're thinking, and in the end delivers a well thought through loyalty blueprint that you can then take and deploy on our core loyalty application.

1:02:31
Awesome.

1:02:32
Well, great, great to have you all here in the flesh in addition to your video.

1:02:36
So welcome to the panel.

1:02:38
And really our theme today is going to talk about a lot of the things that you've seen today already.

1:02:43
These folks are already building a lot of these things and have experiencing some of this stuff.

1:02:47
Not all the new features that you saw, but we're going to talk about a lot of these themes.

1:02:51
So one of the big focuses of this session is really around going from a services firm to more of a product based organization.

1:02:58
So just to show hands in the audience how many people are with services firms in the room?

1:03:04
I would say probably a majority of folks.

1:03:07
Yeah, OK.

1:03:08
And so, you know, our whole panel has actually come from services backgrounds.

1:03:13
And so we're going to talk a bit about their journey from having a services business to moving into products.

1:03:19
So I guess, you know, just open to each of you of the panel, you know, how many of you have or you know, excuse me, what what made you decide to build a product and and move from a services organization maybe?

1:03:31
OK, I'll start with you.

1:03:34
Yeah, one of the reasons is more the the thing that I'm building and that's I'm building an organizational growth platform.

1:03:43
So all my companies are focusing on services and in order to yeah, to create, create new growth, I, I'm identifying, of course new opportunities.

1:03:55
And one of those opportunities is yeah, with a business driven mindset, focusing on building IP, generating new revenue streams and recurring revenue streams.

1:04:06
But the important thing is also when you create new growth parts within your company, it's important to see if you can connect things to each other.

1:04:17
So looking for effects that strengthen each other.

1:04:21
And yeah, but with this new thing with building software, yeah, there's a strengthening effect.

1:04:30
If you look at the Pega behind it, we have experience with it.

1:04:33
And for us, it's helping creating a platform where you can give new people experiences with Pega before they go to the Pega platform, for instance.

1:04:46
Yeah, those kinds of effects help you to create healthy growth company, let's say.

1:04:51
Nice.

1:04:53
And you've decided to actually launch a separate brand for the product that you're building, right?

1:04:57
Yeah.

1:04:57
So I'm decided to six years ago to instead of building 1 company, going for a multi brand company, keep the company small and manageable and focused on their own core business.

1:05:11
So that helps us, Yeah.

1:05:14
Manage the company, let's say.

1:05:15
Yeah.

1:05:17
Great.

1:05:17
Arun.

1:05:17
So you, you've built a successful services business over the last several years in the Peggy ecosystem.

1:05:22
And I know you've built out a team that's dedicated to building product and now you've built the product on Launchpad.

1:05:26
So just talk a bit about why you've done that and and where you see the opportunity is absolutely.

1:05:31
Am I, am I audible to everybody?

1:05:33
I just grab, I'll skip the mic, right.

1:05:34
Grab the mic because we might get you on recording.

1:05:39
We want to record it.

1:05:40
OK, sweet.

1:05:41
So we started Evanses in 2015, right?

1:05:45
We started as a pure play Pega services partner, but we have a long history before that because starting from me, we've been doing Pega for almost like about 24 years.

1:05:56
So the first portion of Pega that I personally worked on is version 3.2.

1:06:01
Steve might probably remember some of the, yeah, some of some of the origins of of Pega as a platform itself.

1:06:10
But as an organization, while we did start as a services, Pureplay services partner, our passion has always been in building products because having build products for Pega in the past, we wanted to kind of get into the methodology of building products because that's where our passion was, right?

1:06:26
So that's essentially why we started a product based organization where we do have services, we also do have products which in tried to deliver meaningful outcomes both from a services standpoint as well as from a product standpoint.

1:06:44
Awesome, great.

1:06:45
And and Brad with the yellow, I know the DNA and kind of original story of the yellow was a services firm and then you folks, you know, built out a loyalty platform primarily on the Salesforce platform.

1:06:56
So maybe talk a little bit about the journey that you've had in the last couple of years with us deciding to, you know, basically take what you built on Salesforce and rebuild it on Launchpad.

1:07:06
Yeah, so our application is our legacy application is a Salesforce native application, which comes with some benefits.

1:07:13
We we get access to the full Salesforce org.

1:07:16
It comes with some challenges where we put a lot of extra pressure on our our customers IT folks because it is a a Salesforce app in their environment.

1:07:27
And so really what we're trying to do is, is, is pull that away, provide the same features, functionality, but also grow our Tam, you know, our addressable market.

1:07:37
So you know, our Tam today or historically has been, if, if you're a Salesforce customer and you run loyalty, we're a good fit.

1:07:44
If you're not a Salesforce customer in your loyalty, we've sold a few right and talked to clients into actually deploying Salesforce licenses.

1:07:52
But again, it's, you know, our, our primary move to Pega was around addressing a larger addressable market pull in some of the the pressure that we're putting on our customers it and then really on the innovation side.

1:08:07
Great, awesome.

1:08:07
So Adrian, your story's a little bit different.

1:08:09
So you've built an innovative AI platform, biz AI and you start to leverage Launchpad.

1:08:16
So maybe talk a little bit about biz AI and how you see the role of Launchpad helping you build your business.

1:08:22
Yeah, thanks Jason.

1:08:23
So yeah, our, our business is quite different because we we weren't a services company, we were a software product company from the inception.

1:08:32
But what what inevitably happens is, you know, you go out into the market and folks want to be able to get an outcome as as John and I talk a lot about and getting an outcome, you know, especially in today's world and in the enterprise space is not usually done with one singular technology.

1:08:51
So in the case of Launchpad and Biz AI, there's a real natural partnership because we provide what we call the content intelligence layer.

1:08:59
And then Launchpad provides the deterministic workflow, the application layer, the permissioning and security infrastructure to be able to actually deploy an application.

1:09:09
So we built, you know, we built a few of these now, you know, regulatory testing, which we call Reiki RI, we partnered with compliance consulting advisory firm in the US to build that.

1:09:20
And then recently we launched A complaints application with the same firm where they were looking to actually take that product to market and they want to own and deliver that software application.

1:09:32
Our technology is actually embedded in the product.

1:09:35
So it's part of the workflow and the outcome we're delivering the, in the case of regulatory testing, it has a lot to do with evaluating evidence, which can be, you know, things as simple as documents, but often it's chats, it could be audio calls.

1:09:50
That evidence needs to be understood as part of the workflow to drive automation.

1:09:55
So we combine the two together and that gets an outcome, which is what actually people buy.

1:10:00
Right, Thank you.

1:10:02
So shifting a little bit into AI coding and development, so you know, obviously a huge adoption in the market around AI coding and development and AI app development.

1:10:13
Why not simply build everything with AI coding agents and why build a good launchpad?

1:10:18
And I'll just leave it open to the panel to to jump in.

1:10:24
Let's start again.

1:10:24
Sure, yeah, yeah, of course, alls are relatively new as well for our company, but I think there's really a strength in the in the platform, right in the platform Infinity, but also on the platform launchpad.

1:10:39
And so there's a lot of being taken care of this with spectral governance, security, that kind of stuff.

1:10:46
And I think the organizations need structure as well.

1:10:50
So with five coding, you can speed up a lot of things.

1:10:54
It really helps make you more productive as we just saw in the movie or in the the the in the demo, let's say.

1:11:01
Yeah.

1:11:02
And but I think organizations also need some structure in the have been building software, keep it maintainable, keep it manageable, keep the cost acceptable.

1:11:13
You can imagine in large organizations things can go really well.

1:11:18
And that's something you not really want to do.

1:11:20
You don't want to create a new shadow IT organization.

1:11:24
So I would agree with them, right.

1:11:26
So AI, as long as it is augmented in terms of building is good, but allowing it to build it on its own and releasing it is not the way to go about it because there needs to be a lot of governance around what's getting built, how it's getting tested, how it is actually getting released.

1:11:45
Especially when you look at, you know, how some of the enterprises are leveraging white coding is it'll help to accelerate what you want to get done.

1:11:56
But there is a governance model that is actually structured around that, right?

1:12:01
So even today when Steve was basically demonstrating it is also, it's also about he, Steve knows the functionality inside out.

1:12:09
So he provides that overlay of understanding the code, what could have been generated and having a proper structure around it, right.

1:12:17
Otherwise, you know, you just build code and then you release it.

1:12:20
You don't know what's getting built, how it is actually going to be used and how it is actually going to be impacting the overall build.

1:12:28
And it's not just pieces of code that is going in.

1:12:30
At the end of the day, it still needs to fit into the bigger picture and that's where, you know, you can't just give it to the OK, let's use white coating to generate everything and migrate to production, right?

1:12:40
There should be structure around it.

1:12:44
Yeah, I'll, I'll just add, you know, we when we talk to customers, one one of the things we say is that the shift that's happened in the last few years to having these powerful coding agents that can help you build and develop software has changed the paradigm from, you know, build versus buy to buy versus maintain.

1:13:04
And I think maintain to me that's the keyword.

1:13:07
You know, Steve hit it again, we're going back to him.

1:13:10
But like when you can use the agentic tools to build something in a model driven architecture, you actually have something that can be maintained, handed to people that understand what's going on, whether they want to use AI to maintain it or they want to actually use the application layer to do that.

1:13:27
It just makes it that much more practical.

1:13:29
And for us it's kind of obvious because that's not our core business.

1:13:32
So you know, we're, we're not going to want to build custom apps that then some software engineering team has to go manage, maintain and evolve.

1:13:42
And I think, I think that that paradigm is really what kind of changes the, the environment.

1:13:49
Yeah.

1:13:49
And for us, I mean, we're, we're a loyalty company, but we're, we're essentially a bank.

1:13:53
So we're rewarding our customers, customers for doing some activity.

1:13:58
And when I do that activity, whether it's sales data or training or whatever it is, we're rewarding generally points or credits or discounts.

1:14:05
And so for us, the, the core has to be 100% correct all the time, like your bank statements are generally 100% correct.

1:14:14
Now the the experiences that we're providing our customers, you know, is, is being byte coded.

1:14:19
But you know, when it comes down to the core, but of how is an activity being processed and what that activity is giving in terms of points or credits or discounts, that's where things are locked down into our core and that's Pega.

1:14:31
Awesome.

1:14:33
And so staying with you, Brad, I think it's awesome what you guys have done.

1:14:37
We're using Replit to basically build custom front end on Launchpad and got a chance to see that recently.

1:14:42
So talk a bit about how you decided to use Replit and where you see Replit fitting or vibe coding for the front end versus Launchpad.

1:14:51
Yeah.

1:14:51
So we build, I think in the video you saw that we have a loyalty designer that allows prospects or even our internal users to to talk with a chat bot, which is the the Pega agent.

1:15:02
And in that conversation, we have to find a data model within Pega of all of the questions that we need to answer to, to develop a loyalty program.

1:15:11
And so the the agent within Pega is managing the conversation.

1:15:15
It's managing what columns of data we have collected and which columns of data we need to continue collect.

1:15:20
And so the agent continues to ask questions until the agent either or to tell the customer or prospect either finishes all of the questions or until they get frustrated, they just say finish it for me.

1:15:31
You make recommendations for me and that all of that data then is stored in in Pega.

1:15:36
What we can do with that data then is very quickly migrate that loyalty blueprint that was created into our loyalty program to to launch a quick prototype for our clients.

1:15:47
Very cool, very cool.

1:15:49
Adrian, one of the things you and I often talk about is, you know, what organizations misunderstand about building products and kind of actually getting them sold into enterprises and the challenges that you have with building something with AI and actually getting it sold through procurement.

1:16:05
Can you talk a bit about that?

1:16:06
Yeah, so our at our company, we what we say is we enable regulated enterprises to turn unstructured content into data that drives automation.

1:16:17
The regulated enterprise component is the key piece because that it is its own space with its own set of challenges.

1:16:25
I think what we've seen is that it's, it's really powerful to be able to kind of get something to happen on a screen really quickly and see it.

1:16:39
What ends up mattering though is can you trust and rely on that output in, in production?

1:16:46
And, and so, you know, at our company, one of the big things we've been focused on lately is what we call trust or assurance signals, which is just all about how do you think about you're getting these outputs from the model, but do you even know they're correct?

1:17:01
How do you know what are the signals that are telling you they're they're correct?

1:17:04
Like it, it is amazing that you can put something in and get something out, But if you can't create that trust or assurance layer that lets the security folks, the procurement folks, the folks that are in compliance actually that or validate that the the outputs you're getting are usable, then they don't get used.

1:17:28
And ultimately, you know, it's a cool demo or app, but it doesn't really get anywhere, which I think is why there's so much of this failure.

1:17:37
Let's say that happens from pilot to production.

1:17:40
You know, this is talked about kind of extensively, but it's true.

1:17:43
You know, when you deal with these types of customers, they're doing all kinds of cool things.

1:17:46
They're working in a lab building really interesting features and prototypes.

1:17:51
But then when it comes to, you know, actually being able to scale that thing out to 1000 transactions a day or a million transactions a month or whatever it might be, it falls apart because you know what what what works in say a vibe coded setting isn't necessarily what works at scale if you don't have that signal and trust layer built in.

1:18:11
So that's kind of how we think about it.

1:18:13
So just to add a few points to what Adrian was actually talking about as well, right.

1:18:17
So if it, you know, track EI, the, the product that we have actually built on Launchpad is, is, is for tracking exceptions and some investigations on a for a cross-border payment, OK.

1:18:29
It's already an extremely complex use case and it you know, there's, there's multiple, but not just one bank that's going to be multiple banks involved in the overall transaction, right.

1:18:40
So it is extremely regulated as well.

1:18:41
So essentially, if you're trying to use one of the things that we wanted to that Launchpad helps is what John was actually pointing about in a speech in terms of the ease, right?

1:18:53
So security is handled as a part of the platform.

1:18:55
So for example, if you are looking at SoC 2 certifications and all of that stuff basically comes from the platform as well.

1:19:02
And it is also the ease of deployment, right?

1:19:04
So most of that the platform basically covers already, right?

1:19:08
On top of it, if you are trying to add the AI layer, which is actually provided by the platform, which is also being governed as a part of the platform.

1:19:16
Because all of these are going to be asked by your CSO when you want to talk about your product, which is what Adrian was actually pointing to as well, right?

1:19:24
You build a cool product, but trying to sell the product through procurement and, and to with the CSO is a completely a different ball game, right?

1:19:33
The advantage or at least a step up that you already have is it's being implemented on a proven platform.

1:19:40
So that hurdle is easier to cross.

1:19:44
Then you actually start talking about the outcomes where your fundamental requirement is covered.

1:19:50
Then the cell becomes easy.

1:19:52
That's right.

1:19:54
Yeah.

1:19:54
So if you know anyone in the audience is inspired today from what they've seen and they want to build a product, how would you advise them on, you know, how they should think about building that product leveraging AI coding tools and Launchpad based on the journey that you guys have been through?

1:20:16
OK, so I'll, I'll, I'll go first, right?

1:20:17
So essentially start off with the blueprint, right?

1:20:23
We didn't pay you to say that, but generally we do that, right?

1:20:27
So essentially when you start, start it off, you, you need to 1st have the idea since I've been doing PEGA for, for almost like what, 25 years now.

1:20:36
I'll, I'll say this, the thing that actually interested me and start doing PEGA is the ease of development itself, right?

1:20:44
But if you, if you want the development to be easier, your design has to be spot on.

1:20:50
So spend a lot of time on the design, which is where Blueprint would actually come into play, right?

1:20:55
And as the code is getting implemented, make sure that there is proper governance process where you make sure that you, you, you check the code, see what's getting built, what's needed, how it's actually fitting in.

1:21:08
Because all of that would play into the maintenance angle that we were actually talking about earlier as well that Adrian was pointing to.

1:21:15
Because if you do not understand what is getting generated, there is no way in hell you are maintaining it, right?

1:21:21
So, so those are some of the things that you have to kind of truly think about as you start using Launchpad to build your products, right?

1:21:30
Building itself, Launchpad makes it a lot easier, right?

1:21:34
But you have to understand nitty gritty details in terms of what's getting generated, how it's actually being used, what's you know in the in the old speak where it is being referenced on what is going to be the impact of a set rule so that you guys need to understand.

1:21:51
So those are my two things.

1:21:53
Yeah.

1:21:53
Another way to kind of frame it is, and people probably heard this framing before, but probabilistic versus deterministic that.

1:22:01
So the way, the way I think about that is like where do you need intelligence and where do you need rules?

1:22:06
And, and actually, it's interesting because the line is blurring more and more because these AI models are getting so powerful with reasoning that they can reliably execute some rules sometimes.

1:22:16
But as, as we saw in the demo, you know, anyone who's done an AI powered demo like is sitting there going, OK, you know, is everything going to happen or the tool calls going to work in the right order?

1:22:27
Is everything going to work?

1:22:28
Everything going to kind of spin out the back end and, and usually it does, you know, if it's architected well, but but it doesn't always.

1:22:34
So, you know, I, I now kind of breakdown things that way.

1:22:38
It's like, let's look at this problem.

1:22:40
Can we just use an expression, a decision rule?

1:22:43
Is this something that just needs routing or some concrete logic?

1:22:47
Or is there truly inference required, you know, intelligence to make sense of this?

1:22:53
You know, and in our world, like in the world of unstructured content, because it is highly unstructured, often that's that's a prerequisite, but many parts of the workflow don't require that.

1:23:03
So I think that's a very easy way to kind of breakdown like what's the role of AI and what's the role of workflow or something like Launchpad where you're trying to drive something with governed kind of restricted controlled outcomes.

1:23:19
Awesome.

1:23:20
Yeah, I'll just add on one thing.

1:23:21
I think spending time on the core, which I think is important because I today we have one agent.

1:23:25
I think if we talk next year, we're going to have dozens of agents, you know, working alongside or inside, you know, our our core application there, there are so many things that you could do, but if you don't get the core right, you can get into trouble.

1:23:40
Yeah, I have a bit of a different advice.

1:23:42
It's more focusing on rethink your processes.

1:23:45
So we are all experienced in building excellent processes with backup and I think the the opportunities there to really disrupt your own processes that you created.

1:23:57
So try to think about how to put angina AI in your process to have, for example, be built in the HR application and look at the workflows that we try to make the HR work more attractive for our employees.

1:24:12
So, and see how you can aid them, for instance, in their own HR journey.

1:24:18
I'm not answer, I'm not just looking at the AIS from a development perspective.

1:24:22
Besides that, I'm not a developer as well.

1:24:24
And so I don't have this background, but I think it creates a lot of business opportunities as well.

1:24:31
Great.

1:24:32
Well, we could price it up here for hours, but we, we need to move on to the awards.

1:24:36
I'm sure everybody's interested to see who's who's won.

1:24:39
And so thank you all to the to the panelists.

1:24:41
And you know, I'm sure if folks have questions, you'll be around for a little bit.

1:24:45
So, you know, feel free to engage you guys and thank you so much.

1:24:48
Appreciate it.

1:24:57
Thanks, Brad.

1:25:00
Thanks.

1:25:00
Appreciate it.

1:25:01
Thank you.

1:25:06
All right, who has the remote?

1:25:09
Where's the remote?

1:25:11
Oh, in my pocket, Jeff, you do.

1:25:21
All right, so last year we had our first annual Launchpad Awards.

1:25:26
This is the second year we're doing this.

1:25:27
We've added in a new award for the Acceleration Award, focusing on partnering in creation of value.

1:25:34
And so we're excited to kind of unveil the awards and recognize folks who've been innovating and building over the last year and doing some really exciting stuff with Launchpad.

1:25:43
So let's go get on with it.

1:25:46
So First off, we've got the Impact award.

1:25:49
The impact award is all about how organizations and their solutions built on Launchpad are delivering value for clients and impact in the marketplace.

1:25:58
And so I'm excited to announce the winner is AI for Process.

1:26:05
And if I get Amit and AI for Process is not only built a great application, they've also built training around Launchpad.

1:26:16
They're also doing outsource development around Launchpad.

1:26:18
So Congrats on it.

1:26:19
Thank you so much for all the work that you've done and we appreciate it.

1:26:23
Thank you.

1:26:23
Thank you.

1:26:24
Yeah, we'll do a group photo.

1:26:25
We'll do a photo at a moment.

1:26:26
Stay on up here.

1:26:27
Yeah.

1:26:28
All right.

1:26:29
And the next one is the Velocity award.

1:26:31
And so just as the name says, it's all about time to value and speed.

1:26:36
And so our winner is unfortunately not here, which is Pronto Solutions in the mining industry.

1:26:42
And they quickly took a vision that they had for a solution that ties into their ERP solution for mining and quickly built that out as Tecton OS as a ERP diagnostic solution.

1:26:56
And so if I could get Teddy to come up and accept the reward on behalf of our client and we'll make sure that we get that optional Teddy way to accept on behalf of Pronto.

1:27:09
Well done.

1:27:10
Congrats, Pronto.

1:27:11
Thank you.

1:27:13
Thank you.

1:27:14
And the next one we have is the Transformer award.

1:27:16
And so this is really all about transformation, modernizing a critical business outcome.

1:27:20
And this is exciting because this has been a company that we've been working with for quite a while with Launchpad International.

1:27:27
And unfortunately, they're not here, but we want to recognize Assist L BDF and a Conicom based in Italy.

1:27:34
They have built out a compliance and risk management solution called Compla Do.

1:27:39
And Luca, if I can have you come up, Luca runs our European business for us and meets with them on a regular basis.

1:27:46
So I'm sure you can personally hand it to them.

1:27:49
Thank you Luca, for all the work.

1:27:53
Good job, Luca, Nice job.

1:27:54
Go ahead.

1:27:54
Great.

1:27:59
OK.

1:27:59
I'm delighted to have the opportunity to share the next three awards.

1:28:02
So the next award that we're going to get into is the Elevation Award, which recognizes a provider that is effectively leveraging Launchpad to swiftly launch new capabilities based on their existing IP, whether that is building a new application or enhancing existing products to address evolving client needs and emerging market market opportunities.

1:28:22
And our winner of the Elevation Award this year is Avansys.

1:28:27
So Arun and team would like to come back up building on their years of answer for of experience of financial services and payment investigations.

1:28:35
Avon says Arun mentioned has created a modern solution that helps banks automate payment exception handling and investigation.

1:28:44
So we're going to go, Arun, Congrats.

1:28:46
Do stay on up here with us.

1:28:48
And our next award is the Apex award, which is the Apex of awards and that celebrates the provider who builds a highly sophisticated feature rich solution on Launchpad that incorporates advanced functionalities and demonstrates exceptional technical depth and complexity.

1:29:05
And our winner of the Apex Award this year is Filo.

1:29:12
I think some of the stuff that's really interesting continue to push the boundaries of loyalty management by incorporating AI powered and agentic capabilities into their platform.

1:29:20
And that brand new loyalty builder that Brad mentioned helps organizations design and optimize loyalty programs through AI guided recommendations, which accelerates speed to value while improving their customer engagement.

1:29:31
Brad, congratulations to you and the team.

1:29:35
And finally we have the inaugural acceleration award and you know we're on our second annual awards ceremony.

1:29:44
So why not inaugurate a whole new award in round two.

1:29:47
And this year that award is for partners that are helping Launchpad and our broader Launchpad community accelerate value creation, whether that's through services, through technology or through to through go to market support.

1:30:00
So the first partner to receive our acceleration award is AWS and I'm gonna invite Greg up to recognize to receive that award.

1:30:10
For folks that don't know, Launchpad is on the AWS Marketplace, making it easy for new customers to sign up and build products of value on Launchpad.

1:30:18
Launchpad providers are also accelerating their path to the AWS Marketplace as a distribution channel using Launchpad's full technical review to help them accelerate the AWS Marketplace onboarding process.

1:30:30
But one thing that I'm also really excited to announce is that we have just launched a new offer with AWS for startups in which Launchpad is matching AWS incentives for startups.

1:30:40
So we are jointly supporting startups efforts to build and deliver outcomes as a service solution.

1:30:46
So excited to work with you guys on the AWS side.

1:30:49
Thanks, Greg.

1:30:54
So congratulations again to all of our Launchpad award winners, and my thanks to everyone who lets us work with you to build your products and grow your businesses.

1:31:02
I hope that we'll have the chance to work with everyone in this room in the coming year on the delivery of outcomes as a service.

1:31:09
And ladies and gentlemen, that wraps our Launchpad session.

1:31:12
I hope you have a great rest of Pega World 2026.

1:31:17
Thank you.

1:31:23
We all have a photo.