PegaWorld 2025: Capgemini AI-Powered Legacy Transformation: Break Free from Your Legacy Apps and Fast-Track Your Pega Infinity Journey

PegaWorld 2025: Capgemini – AI-Powered Legacy Transformation: Break Free from Your Legacy

Good afternoon everyone. How was your day? I'm Dinesh Karanam. I'm from Capgemini and I'm with my colleague Nicholas, who is who is part of the complete solution. He will walk you guys through all the details of it. And of course, with Matt.

You guys are already aware of it. And he will be part of the initial process of it. And now with that, I will hand it over to Matt, who will walk us through a few of the slides and then we'll go to that. Awesome. You guys having a good PegaWorld? Yeah.

I apologize if my voice cracks or anything like that. I was screaming at the concert last night. Um, I don't know. Mark McGrath brought something out of me. It was a good time.

Um, so I'm excited for this session because yesterday on stage, you saw Kerim showcase how Blueprint can take in more and more legacy system assets to inform the future state workflows and really sort of pave the way to the cloud. But of course, you don't always have a lot of those assets, which he referenced, laying around comprehensive documentation of your applications and understanding of those legacy applications and what they do. So we've been working with Capgemini over the past couple of months to really integrate some of the modernization approaches and tooling, which they already have with our Blueprint led approach, to allow you to quickly understand your legacy systems and then go through that blueprint process to create new workflows in the cloud that enable you to shut off those legacy systems in the end. So I think we all know why it's imperative for us to get off of legacy systems right now. Um, legacy systems, you know, they incur so much, uh, cost, which is right in front of you in the form of maintenance costs and operational costs.

Right. You spend so much like bringing in consultants to make sure that your COBOL systems continue to work. Make sure that your Java applications don't break. Um, making sure that, you know, new features that you're releasing are properly integrated into these legacy systems. So there's a line item for legacy system maintenance, but there's also a hidden cost and opportunity cost with legacy systems in what you're unable to achieve in terms of customer experience gains and in terms of automation, which you can deliver to the business.

So we're really focused on helping you get off of these legacy systems and do it quickly, paving the way to the cloud. And I think I'm going to turn that over to you to show me how. Thank you, Matt, for that introduction. So as Capgemini, we have a lot of smart people and some of the smart people work for our research institutes and the research institute publish a lot of reports every year around industry trends and technology trends. And one of my favorite reports recently is around GenAI for software engineering.

So I completely I read it because I'm a complete geek and I love to understand the latest trends in software engineering. And one of the kind of figure that's caught my attention in that report was the one you see on the screen right now. And it's really outlines how the different paradigms of programing languages and Platform has changed the way we develop software and delivered value for our business. So like back in the Stone age, we did like assembler language. I don't know if anyone in the room actually knows that anymore, maybe a few of you, but after that we went on and kind of abstracted away from that, and we had this cool COBOL language that was supposed to abstract away and be the low code version of the 60s and 70s.

Of course, we continued the involvement, and now we have object oriented programing and we moved on. And now we work with Pega. We know low code, cloud native like we known it for years, so we know the value of it. And also if you listen to the last year, you know the value of low code development on Pega, it's eight times faster than Java. So yes, that jump wow.

It's really tremendous. Um, but now like in the recent two years, I don't think anyone has missed GenAI. Right. ChatGPT we saw it on stage this morning again. And the hottest new programing language is English.

So quite cool right? We do live coding now. We're vibing. Yeah, yeah we're vibing away. But like we're still stuck with these old apps. So how do you do vibe coding if you're stuck with COBOL.

So and this is not a challenge that's new for for for many of you, I guess since you're here in the room, You know that this is a challenge you have, but at the same time, you cannot really build a business case around it to make the transformation happen. Because like, it's a lot of effort building, like the knowledge to kind of start this journey, getting the insights of what your old application did or still do. Um, and also the benefits of doing this might be a very long journey. So again, it's tricky. Um, and it's very cumbersome to start this journey.

So should we give up? I don't know. I don't think so. Because now we have, of course, GenAI and a lot of cool tooling. So GenAI is really a game changer for this. So how do we apply GenAI and this type of scenario.

So first of all we can do it use it for jumpstarting discovery. We also Blueprint demos of the past days. And hopefully you tried it yourself in Innovation Hub. So you can really use GenAI power of GenAI to design your next application. We've all seen the power of this.

We also can you can use it to optimize workflows to understand what are bottlenecks, how can we improve our applications as well? And the thing I'm very excited about now is also to set the foundation to finally retire those legacy apps. So this is a real game changer. What does this actually means? That we're changing the equation. We're changing the balance. And now actually we can reduce the effort to analyze these old apps and also realizing value much quicker.

So it's a real game changer for us. Um, so many of you might, might kind of recognize this scenario. You have your old legacy architecture. You have your UI and rules kind of intertwined with each other. You have hard coded processes that are very rigid, so you cannot really keep keep up with the change that your business requires.

Um, and, and you have like these monoliths that are really hard to kind of break apart. And also you might have like stored procedures and whatnot in your database. So it's like your data is trapped and the logic is trapped in this database or whatnot. Um, so it's um, it's really tough, but just keep calm. So calm is a solution and service offering.

We have from Capgemini and the Calm abbreviation, which is not only easy to remember, it also stands for something. It stands for Capgemini AI Assisted Legacy modernization. So what this enables us to do is we have done this for for many, many years, many decades, modernizing applications. But I'm happy now to bring it to the Pega ecosystem. It enables you to understand your complete application landscape, building a business case and trying to really understand where does it make sense? What should I do with my application portfolio? How can I save costs? How can I be more agile so we can start already there? If you already have a knowledge of like, this is the application I want to address, or you get that from the landscape assessment, we go in and do a deep analysis of one of your applications or several of them using our tooling of course. As part of this, we generate a lot of documentation that gives you the insights you need to transform your legacy apps. We also do the gap discovery because we don't just take what is. We also make sure that we understand how can we enrich this and make it better. What do we don't know still. By analyzing the code, what is the knowledge we actually need to have a workshop around rather than going workshopping it all away.

And then we bring this into Pega make this cool app on Pega leveraging Pega, GenAI, Blueprint, and also bring the data along so we can also do have tooling to to migrate the data and make sure the data integrity is there. We can verify the application. We have also a lot of good testing tools like test maker to do automated testing in the Pega ecosystem. So this is the kind of technology part of it. But we are not just implementing Pega or doing a migration.

We're also doing the strategy and the change management all along to make sure that on day one, your business is ready for the change. So it's quite cool and I'm quite excited about this. Let's let's go into a little bit more of the details of how this actually work. So to enable Calm we have something we call the Legacy Modernization Workbench, which is a collection of battle tested tools from Capgemini. These tools have been used already in production for many clients to analyze legacy applications.

What we have done together with Pega is making sure that these tools work together with Blueprint. So we can now go from this old COBOL interface using, for example, Cb360. I will tell you how soon and generate a lot of different assets that can be consumed by Pega and Pega Blueprint Other. So digging even deeper, like you have these old codes based on, for example, Lotus Notes, C-sharp, Visual Basic, COBOL, whatnot. Okay.

How do you do that? Then you import it into our workbench where actually we let our workbench look at your code. Okay. And then we can generate a lot of these assets. So we generate an application overview. We generate an application model a complete model of your source code and all the insights we can gain from it.

So. So you can actually then query the knowledge we extract from your code. We generate personas. And I think you see where this is going. We generated personas that uses your application today.

We can also generate workflow models in the BPMN format that Don hates. but they are good for reference and also you know how to use them in Blueprint, right? We generate workflow descriptions, so you get for each of the workflow we identify in your old application, we get a high level description of what it does. And then we also can understand which personas are using this and which portal is it in, started and so forth. We extracted data models. So if you use Blueprint, which I still hope you have done, we we can.

Now when you do a Blueprint, you see the case flow and you see the case data. And we follow the same principle here. So we generate a report of all the data fields we found and on which workflow they are used. But we also generate something called the DDL file which is a representation of your data structure. Um, we are also extracting all the business rules and get a full library of the business rules in your application.

And not only that, we track and trace, where are the business rule actually use and which process? And so we can actually see if there are business rules used in multiple workflows along your application. We also get an insight of the data dependencies. This is especially important when you go from mainframe because like reading mainframe code and understanding like what's happening and tracing that it's it's tricky. And getting help to analyze and understand how to extract all the data and make sure it follows along in my new cloud native application. That is something we can use with our data dependency reports.

And also we make a first draft of the system components that you should do in your new modular Pega applications. So it's trying to decouple this monolith using our tooling as well. So this is really powerful. And as I said the tooling already works before we talk to Matt about it, but I met Matt last year at PegaWorld and like, you should really look into doing something with Blueprint with this. And that's when we started to kind of make sure that these assets were tailor made for Pega Blueprint.

So what do we do with these? Um, these all these assets we, uh, we can from this, derive a few patterns on how to migrate your old applications. Um, and I think already last year, you could start to import BPM and files from your workflow based applications. And also during the year it was added DDL import and whatnot. So already if you had like IBM, BPM or Appian or any older workflow based software, you could extract the BPM files from those applications, import it into Blueprint. And what we did with IBM at Capgemini is we took this, infused it with our industry knowledge when we worked with our clients and made sure we built a new app using Blueprint.

We did this for many clients and it was really powerful. Just doing that bit. Um, so but that was kind of for the workflow based application where it's clear that this is a documented workflows. We have the BPMN files or a system that can export it. But now what we can do is we can target those custom built apps that still are somehow workflow based.

Maybe it's not explicit in all cases, but they are. There are workflows that the users actually follow. There are rules in the applications, and there is a lot of logic that you need to derive. So using this workbench and extracting all these assets, we can now import them into Pega GenAI Blueprint. You saw this working on stage yesterday with Kerim as well, and it's the exact same procedure.

Now we enable that legacy input app part that Kareem demoed on stage. This is what enables you to do. And the third pattern I still think is important to mention is if you have all these data in your app, because the data is an important part of all the applications. So here you have to understand if there is any logic trapped in the databases where you have to extract them and lift them into the application layer. And then we make sure that data follows along and is migrated.

And we verify that all the data came there with our experience on how to do these migrations from on prem mainframe or whatever it is, into a cloud database. Right. So if you have a custom application based on code, like how do you do it? Like is it overwhelming? Yes. If you do it by hand you probably can't. But I will give you an example of one of the toolings from from our workbench or two Toolings, actually.

Um, it's called Gen Insights. Um, so what Gen Insights can do is, is two purposes mainly. In this scenario you can take the legacy code and just do a quick scan. Do a profiling. What is the nature of this app.

What was the technology. How complex is it. Like how big is it? All these things? You get a complete profile of many of your applications, because this quick scan enables you to scan a multitude of apps very quickly, and you don't even have to share the source code with us. We can run it in your environment just using our insights plugin. When we get the insights of all your application landscape, we can import it into a tool we call Clearsight. What Clearsight enables you to do is this first step I explained in the process the application landscape assessment, and I will shortly show a video of the whole end to end flow. But it's really enables you to look at your application landscape and understand more, gain insights and also recommendations on what to do with it. Why is this important? Because it helps you to focus on the right things, so you can build a business case for the legacy transformation and also build a roadmap. You might have multiple apps you need to transform. Where do you start.

And this is something Clearsight helps you with. And then once you have identified the candidates and the roadmap for it, we again use any insights or any of the other tooling if it's more appropriate to do a deep scan of the specific application and create these assets and use in Blueprint. What if you have really, really old app like mainframe? This could be my grandma actually looking at the mainframe application there. And what do we do then? So I worked with also our legacy modernization team within Capgemini who look at mainframe modernization. And we have, again, battle tested tooling that's been used for a long while to analyze COBOL code or mainframe application, and understanding how to break them apart, to be able to transform them into something better and is better is, in many cases, Pega now.

So what we do is we take all these COBOL code, the Copybooks, and all the data dependencies. Let Cap3 60 analyze it. Cap3 60 can then be used to generate like an overview of your application. But we can also then use something we call bread business rules extractor and Documenter. And that enables you to do is extract business rules from COBOL code and explain it in English.

The new hot language that we're all programing in now, and together with our expertise and these tooling, we can start to decouple it and understand the dependencies in that old mainframe app and prepare it for migration. So we basically make logical divisions of the application, and then we can import it into insights and generate all these assets that you need to make a blueprint of your new application. So this enables you to to kind of reduce risk in the kind of preparation of the migration and actually the migration later, but also make you it makes you able to make it faster to make this analysis. So with this said, I want to kind of bring it a little bit more to life. So time for a lift off.

I want to show you a demo video, which I will explain what happens, like end to end, moving from this whole big landscape, which you don't really understand what to do with into a Pega application, end to end. Oops. Right. Right. So what you will see is how we go from this whole application landscape assessment and follow through all the five steps that are outlined in the beginning of the presentation.

So we start with looking at Clear Site, the IT decision Maker tool to analyze your application landscape. You can look at your application landscape from a lot of different angles complexity, number of tickets, what type of source language it is like. Is it old technology that's actually out of support understanding even CO2 consumption in your hosting of the application and really understand like from a different angles. What do I actually have? You also get this checklist and preparing for your transformation by checking. Do I know everything I need to know to start a transformation project? To get support and answering the critical questions to make a successful transformation of your application landscape? So you have to prepare for that.

And also we have tooling to kind of help you select the best candidate based on different business rules. Like this is an application based on this technology which is out of support maybe, and it's a workflow based application and so forth. And hence it's actually a good target for Pega based migration. So it really helps you build this roadmap and you get these recommendations on what should I do with my application landscape. What is the different modernization strategies I should use? We then move into the application analysis part where we do a deep dive of your specific applications in the order we decided with it.

Decision maker. And here you recognize the kind of slide that's animating. So we go from these old apps. As I said, we can do Lotus Notes, C sharp and whatnot. Um, and analyze it very forwardly with, with our tooling.

So we run that on the code and generate a lot of documentation. So we generate this, as I said, application overview, the business model, the business process and all the data models. So once we get that you know where to go, we do have a look at the documentation of everything that was analyzed in the source code. So here on the screen we see how Threesixty enables you to get an overview of your old COBOL applications and dependencies. And what Copybooks do I have here.

How do they relate to one another. So you really get this dependency which helps you understand how to decouple all the dependencies in this monolith. And also you have access to data dependencies report already in this stage we generate this application summary. So it's a high level description of your application with all the workflows, the personas and everything. It's a good good for takeaway anyway, but also it's a summary of your application.

We generate the workflows, the fields that are in application and so forth. What we then do is we go into the discovery phase. So what we discovered when we generated these assets based on on insights and all this application overview and whatnot, we realized we're throwing away a lot of good knowledge here because summarizing a big app in two pages, like there is a lot of information lost somewhere, or at least it's good to have at some point. Um, so what we're doing in this stage is actually we're creating a knowledge graph of all the, the, the knowledge we generated out of your analysis, and it enables you to have this kind of transformation assistant, which you can then query like, what are my most like most referenced business rules? What what is this workflow, how often is this data used and so forth. So you can actually use this to prepare for your migration but also throughout your migration.

So it's a really powerful friend to have along the journey. What we then do is like we're kickstarting the implementation journey. So we're going to Pega Blueprint. Um, and here we are really leveraging and taking all these assets, uploading it. You saw it on on screen.

I didn't have access to the legacy import that Kerim had yesterday. But once we get access to it and you get access to it, um, you can actually take also a lot of the code insights uploaded into to Pega Blueprint. So we upload it there and let Pega do the magic of generating and analyzing. What you also see here is it's Capgemini branded Blueprint. So it's not a bug, it's actually a feature.

So what Capgemini have as one of the few partners, I'm happy to announce that we have something called a partner powered Blueprint. And what? This is different from the regular Blueprint, which is still amazing, but the partner powered one from Capgemini is infused with all this industry expertise and knowledge best practices from our previous projects, which helps you to kind of not take just what as is, but we infuse it with our best practices, best solutions, everything. So you get it as part of of the partner power Blueprint. And as you saw, I just uploaded a document and it was pre-populated. It even understood in which sector the application was.

I get all the workflows generated for me and I get an overview of those. I get a detailed workflow, you know the drill and at this stage I get the option also then to upload the BPMN file if I want. Or I can go with the one that was generated through the description of the workflow. Both options are available depending on what you think works best for you. And then in the next step we look at the data and integrations.

So in case of we identify integrations we get the definition of the interfaces. And in this case you see we're uploading the DDL file and capturing all that data that we found as part of the analysis. So not only do we capture what is needed to run the workflows that are generated, we also uploading the legacy and capturing the legacy data that you actually need to bring along. We capture the personas and as you know, you can also at this stage define this kind of security structure of the application to understand who can do what, launch which cases, read what data and all that. And then we get this summary and also the favorite the preview of your application.

This is very powerful to kind of make it come to life even before you implemented a single line of rule in App Studio. And also, as you saw on stage this morning, we get Agentic for free. In a sense it's quite cool. Like, look, you can go from a COBOL app into an Agentic app this easy. It's quite amazing and it's mind blowing if you think about it, like how you take these old looking applications and get an agent as part of your implementation.

Um, so once the blueprint is ready, we import this into the Pega environment. So it's really time for liftoff soon. Here. Um, so once you import it into Blueprint, you get your foundation for your application ready. All the data is there, the personas is there.

Um, and you really kind of reduce the time dramatically to implement your new Pega application. And, um, after this, of course, we iterate, build the MLPs, and then finally your business is happily using a new business experience with a genetic AI automation and not being stuck with this old legacy application. So that was like a quick end to end journey of how we can do it together. Cool. Um, if you're interested in actually seeing more details or kind of understanding how this would work for you, um, we would be happy to kind of schedule a detailed demo or looking at a specific application for you.

So making sure we can use GenAI for you and doing pilots together with you and really kind of come with with a solution that helps you break free from those legacy applications for once, once and for all. What I really like about this is, you know, at first you you kind of ran me through all these tools and I was like, bread, clear sight, jet AI inside. I was like, what's going on here? But then I realized, like, it really does help you create almost like a portfolio funnel of a top level understanding across multiple legacy applications. What's going on at your estate? Then start to dive in and understand within each of those applications comprehensively, what do they do? And then summarize it so you can quickly get into a blueprint to start to crank through actually getting those applications out of legacy and into the cloud. Nice summary and good to get an outside view.

Thank you Matt. Um, so so please do tell if you you you want to know more and have want to have a pilot on your specific app. We as I said, we we can do a multitude of languages and make sure we get on to Pega. I'm really excited about this. With this said, I want to hand over to you, Dinesh.

Thank you. Is that good? We just learned new. Okay, so apart from what you have seen, the industry domain knowledge as well as every Blueprint, whatever we created, it's a branded Blueprint where we will bring our own assets, where you will see the data model as well as the Constellation components. You may heard many times there is a Rd components where it comes up. All those things comes along with it as part of the overall.

But definitely we have to do some customizations for it. Then end of the day, what you can see here, what benefits you will get it, the faster way to go to market strategy of it, and quickly coming out of existing legacy applications to the new platforms. The third one is again, you're making sure the entire the rules, what is buried inside the COBOL code or the legacy applications. Bringing those into the natural language of it, which is very difficult even if you use the entire COBOL code and saying that, okay, tell me what is the logic is written. It's not going to be that easy to bring that.

That is what exactly the power of com Ofcom. That exactly will tell you what methods has been written, how these methods are reused across various platforms or various components of it. Third one the data model. You can automatically see it OK. It will bring the existing data model and also from Kara perspective to the domain models.

What we have it. You can compare and see what need to be. The new changes need to be done for the new application. Okay, so with that, that is what the overall cap. Now I'll talk about it.

Who we are. You guys may have seen various things about the cap. We are not a small. It's close to billions of dollars, right? So it's a huge company, close to 370,000 people across the globe. And we represent almost every country, wherever we have it.

And we have a large pool of Pega resources who are certified resources. Anytime we bring in the big differentiator we always bring in. So these are the various countries where we represent and implementations, as well as the delivery centers where we have it. And this is what our overall Pega we are working with Pega close to 20 plus years and last 20 years we were every year we were recognized for every year as a partner of it. Okay.

And we worked with very closely with Matt and the rest of the teams to make sure what is the industry is looking for, how we can able to improve or bring the new frameworks or new features of the solutions for it. Okay. With that, again, I want to talk about it at the last one like a business value. I don't want to go through every number here. These are the various numbers what you can see.

These are the values where we work with UN as well to bring those and working with them the sustainability and all other aspects of it. Okay. Then what makes Capgemini different. So you can see one is the domain. The second one is the tooling set.

What? You looked into all the tool sets. These are all battle battle tested. So every tool has been implemented more than a decade or so. Whatever you can say the cap 360 or bread framework are genuine sets is close to say three four years. So we are building and we are keep on enhancing as and when there is an applications coming up, what new features need to be? What new technologies are coming up? Okay.

So if you have any questions again, as we spoke about it and you wanted to do test and test some of the small applications or any of your complex scenarios, bring it to us. Definitely reach out to me or Nicholas and we are happy to hear help answer, and we'll work with you to bring some of the new things, and we'll test it through the application. What we have shown the complete framework. Okay. Thank you very much.

Thanks..