PegaWorld 2026: AI Unlocked: Legacy Transformation by Excellus BlueCross BlueShield

Good morning, everyone, and welcome to the session on Legacy Transformation case study by Excellus Blue Cross Blue Shield. I'm Dev Victorien. I'm the Director of Intelligent Platform Services from Cognizant. I would like to introduce our speakers for today. Todd Milillo Director of Business Transformation. And we have Len O'Neil our solution architect from Excellus BlueCross BlueShield. So how many of you are have done Lotus Notes to Pega migration or currently doing a legacy migration to Pega. Anyone in the audience? Okay, so today's agenda. We are going to tour and then are going to talk about what are the key drivers that really push them to move away from Lotus Notes to Pega? Like what are the key business drivers? Right. We're going to talk about that. And what are the various methodologies that they took in order to do a successful implementation? And then also they would talk about the solution.

Architecture was a high level solution architecture that was created for this Legacy two Pega modernization. And also we would cover the outcomes and impacts and some testimonials. Towards the end of the presentation, I'm going to cover about a very key solution accelerator from Cognizant, an AI tool, which is going to enable us in the speed to execution for projects, for transformation programs from Legacy to Pega. OK. We'll cover about that as well. So let me set the context about Cognizant experience in this the space, right? As you all know. Cognizant is one of the oldest and the largest Pega practice in the world, and we have a huge certified skill talent of Pega associates with us. And we have enormous experience in implementing enterprise grade, like complex, huge transformation programs across the globe. Okay.

And then, um, the biggest differentiator between Cognizant is not just about the scale, but the key focus on like the value that we bring to the table with our extensive domain and technology experience for the last 30 plus years in the industry, and also primary focus on innovation on a day to day basis, I think that is our biggest value differentiator, I would say, and the innovation in terms of low code and then intelligent automation, as well as AI led innovation that really helps us to partner with great clients like Excellus BlueCross BlueShield to achieve like bigger transformation programs with ease. And Todd will cover more about that in the slides. But now I will hand it over to Todd to present their transformation story. Thanks, Deb. So I think we have a mistake because I heard Little John was supposed to be introducing me today, and I don't see him here yet, so.

Um, so thanks for the introduction. We're from Excellus BlueCross BlueShield, where the largest nonprofit health plan in the central New York area, covering 31 counties, over 4500 employees. And it's always important to start off with our mission. So it's to help provide help people in our communities, live healthier, live healthier and more secure through access to high quality, affordable health care. That's something that we drive everything through with our mission, and it goes through every project that we that we're involved in. So when we started this project off the modernization reimagining, we had a Legacy Lotus Notes platform. And I'm assuming not a lot of hands went up, but people who were on Lotus Notes. But I'm assuming there's legacy platforms on every single one of your organizations. And it starts off with the challenges. We didn't just jump right to Pega. We said, what are we trying to solve?

And looking at that is inside there. We had no way to report. It was hard to maintain our business was not getting any efficiencies, and we weren't even sure what was happening inside of Lotus Notes at the level we needed to be to be efficient. So looking at all those challenges, it was time to modernize and modernize. Wanted to solve the challenges, but we also wanted to put a foundation that we could build on and not just do like for like, but it was a chance to redo it and think differently. So we went from Lotus Notes to Pega. And it's pretty obvious because we're presenting at PegaWorld. So during that, we worked with Cognizant and we started our journey into replacing our Lotus Notes? Platform and looking at ways we can impact our business in a positive way. And one of the keys that we had, and we've used this model for a lot of our initiatives, and it's been very successful working with Cognizant as a partner.

It's shown that this model has allowed us to be on budget, on time, and produce quality outputs of our of the work we're doing. And it starts with discovery. And the discovery phase is something that we've done now for all of our major initiatives. It takes about eight weeks. And you start thinking like, hey, two months before you start anything, that's a big investment that investments pay dividends in the end. We take during discovery Cognizant and our Excel as partners, our SMEs all get together and have these sessions for eight weeks straight. And we talk about what is the business trying to do. They'll bounce ideas from other areas. So it's not done in a silo. Other areas of the business will come in and say, well, I need it to do this. If you do that and it starts out of that, we end up with high level requirements.

Everyone comes together and reads through all those high level requirements, make sure we don't have any gap. But now we're not done. We understand from there the size of the project, we understand the scope, gives us a really good estimate on cost and what it's going to take to do it. Next we have our DCO sessions. Those are the deeper dive requirements. Still the same concept. We bring everybody to the table. And as we have these conversations and drive down into the how it's going to work, we end up with fully fleshed out epics and a project plan that will drive us to success. And then working in an agile methodology, we have our demos, people see it give feedback, and what we found is by taking all that upfront time, there's a lot less changes, a lot less rework happening during those demos. So a lot smoother, smoother interactions at the end.

Not to say it doesn't happen. We still have things come up, but it has drastically reduced the way we implement our work. So deliver it now. Next we got to do the adoption. So we deliver it out into production. We have our training, we update our documentation. And when we went to this place from Lotus Notes, there was a little anxiety about switching platforms. Lotus Notes, it's been there for a long time. We're an older organization. People have been there using this for a very long time. We're replacing it. What we found, though, is obviously some resistance to something new. But in general, once people started using it, that resistance went away and it started working. It was more modern and now it's just part of our day to day workflows. Then I'm going to turn it to Dylan. All right. Thank you, sir. All right.

I'll be a little bit more podium present here. So, um, one of the first things in my role I have to acknowledge, as we were talking about it last night, and my role, obviously going from conceptual design to realization of the implementation with this team has been amazing. Like I said, just watching it like how our projects move from that conceptual design and realization. Great partners. So now let's talk a little bit about our approach to the solution architecture as we move through this transformation. So basically the first thing we want to do is like basically apply some guiding principles on, which basically evolved from our enterprise principles, from our enterprise architecture. But these were key in helping us like design the solution as we move in. Migrated from Lotus Notes to Pega. The first was a real time integration first strategy, right? You know, I'm sure a lot of us probably still, you know, have a lot of batch process.

But one of the things as part of our modernization, we want to make sure that we like kind of, you know, modernize our data flow as well, which makes us more responsive, makes our systems better, easier to integrate. So with that, we obviously prioritized APIs first as well as event driven architecture. Oops. There I go. There we go. All right. Thank you. Event driven architecture. So APIs. So basically our API is basically enabled us to be able to integrate our systems real time with these new workflow applications. Lotus notes right. We really didn't have the ability to do API integration. So members or users would, in many cases, would have to navigate multiple applications to pull the information they need for their notes. Lotus notes case to populate it or, you know, actually or even push information. There was a lot of swivel chair. Those APIs enabled us to do such things as integrate with our primary systems, which we'll talk about, um, and pull that information in.

For example, I'll share a good API example. Shortly. EDA. Event driven architecture. This is something that's relatively new to our enterprise, but it's something that we, you know, making every effort to kind of, um, implement across our system, across, across all of our applications, which basically enables like real time responsiveness and real time data updates within applications. One of those four examples, for example, are our is using CDC in our Oracle databases, right to capture real time changes, publishing them through our Kafka stream, which Kafka really is our backbone of our EDA infrastructure. The second principle minimizing customization, um, we made, we intentionally made sure to minimize any custom, um, custom code within Pega. Um, which basically allows us to reduce long term maintenance overhead, enabled faster upgrades and stay aligned with the Out-of-box Pega capabilities. Basically, this, you know, making sure that we ensure that our applications stay sustainable and scalable, not just like quick fixes.

Third principle build once use everywhere. So here this is where we kind of embrace that. Pega. Situational Layer Cake. We basically use it to create a common layer that we can share across all of our applications, even across our multiple instances, to make sure we centralize reusable logic and components, eliminate duplication across applications, promote consistency across workflows. Basically, it accelerates delivery of each application. Um. Excuse me. And as you'll hear shortly, this strategy played a key role in enabling us to better leverage our enterprise platforms. So as we migrated our applications on Pega workflow, we unlocked true integration. So I talked a little bit about that API integration, which basically enabled us to, you know, affect in the end automation in some of our applications. So systems such as our claim processing systems, where we are able to now pre-populate claim information in the application workflow, our member systems, we're able to pull or sometimes push eligibility information to those, um, core systems.

So there's several systems provider I could go on care management system. Each of these were key integrations, which previously within Lotus Notes there was a lot of swivel chair. So so the key benefits realized from this is the elimination of that swivel chair, right? Navigate multiple applications, copy and pasting information from one sometimes, um, copying and pasting duplicate information. Um, obviously you can see that's prone to that, can be prone to data errors or copy and paste errors. So reduce data entry errors is another key benefit that comes directly from this integration. So basically it ultimately results in improved user experience and obviously more efficiency and less data errors as we talked about. So one of the other key benefits is that right. We just talked about swivel chair. Now there's that one unified interface that users have to interact with when, um, managing their flow or even submitting their flow.

So those users no longer have to navigate to, to the multiple systems. Which result in significant productivity gains and a better user experience. All right, so we talked about the integration. So here's, you know, with this integration, the API and integration has allowed us to do. It basically enabled us to better leverage our enterprise platforms. Now we're setting up a, you know, a lot of new enterprise platforms, such as we talked about EDA. We also kind of setting up a customer communication management system, right? So for example, with those systems, we're able to build that common layer that we talked about in the Situational Layer Cake and duplicate it across Pega applications. So for example, right. Each application no longer has to have its own unique implementation with the customer communication management system, right? We build that once and deploy it across all applications. So now we have one centralized set of code set of capabilities that we need to manage.

And then each application now can integrate and interact with our customer communication management consistently. This same principle was used across our content management systems. Again, you know, something new. Where now? Right. We have a single file repository document where every application have the same access interact in the same way. Our customer profile system, another system that was key where now we're able to, you know, number one, provide a unified view of customer member information within the application as well as be able to obtain the preferences right of each individual member. So for example, with the communication system we just talked about. Now within that workflow, when you operate within the context of a customer, we can send a communication out using that, but according to their preferred communication method or channel. So lastly, what migrating these applications to Pega workflow basically enabled us to have tighter integration with our CRM application, right?

So this is right where our customer service advocates work as their service members. As you can imagine, our customer service advocates previously having to navigate to other systems like say, for instance, open a advocacy advocacy case, right? Where previously they would have to follow a process that navigates to another system, copies data again and, you know, submits that. Now with this Pega workflow integration, we can use our API integration, specifically DCS API. The API has been huge for us, which allows that, you know, inter Pega application integration, as well as even exposing that in Pega integration to our external applications, allowing them to also submit cases. So now within, you know, many of these applications, we can simplify the experience of opening cases and even viewing the results or status of a case. So imagine a member calling in and you know, what's the status of my advocacy case, right in CRM.

We're now able to present, you know, a details of that status and then potentially even click navigate right into the application right from the CRM to get more detail. But ultimately that integration really, um, lends, and it extends our mission for presenting the 360 degree view within our CRM. So in summary, our modernization with Pega has driven real time, excuse me, real time event driven integrations using EDA and APIs, reduce customization and improve maintainability. Reusable architecture utilizing utilizing a Situational Layer Cake model, full workflow automation through deep, deep system integration and elimination of swivel chair inefficiencies. So with that, thank you. Thanks, Ellen. Just a little bit more. And this is a adoption and change management and how we work together between both organizations. And this is really important because a lot of times you'll work with multiple people from other organizations and it feels like a divide.

So when we worked with Cognizant, it really felt like a partnership and we didn't have meetings of us, meetings with them. It was one group. Everyone invited full transparency, but we also had our specialties. So as you see, we kind of broke apart where we were. We own the architecture, enterprise architecture. We developed it. Here's our vision, here's what we're trying to do. But then Cognizant helped bring some of their thought process to help do the build. Engineering brought new ideas to the table. We approved all the all the stories, the priorities, made sure we accept them into our environment. The testing, something that we're really proud of is our UAT testing that we're doing across all our platforms. We have our our business involved writing test plans, executing test cases, making sure that we can accept this into our environment. It's just another piece of the puzzle of how we ensure when we transform, when we add these new capabilities into our production environment, they're able to be used by the business and the way they expect.

So that really robust UAT process has helped us at the end, and it gets us to where we want to go. So outcomes, we're looking at this and we're seeing the outcomes. But unfortunately, due to our Lotus notes it's hard to measure where we came from. So the lack of reporting is really hard to tell the full story. But we do know we are at 10% additional cost savings for test automation. We have some metrics around that that's able to measure. Like Len mentioned, we wanted less customization, so less than 5% customization for our implementation, which is really going to help us not have to maintain custom code throughout the future. Our process improvements, we're seeing a lot of efficiency. We're hearing it from from the people using the tools. It was a huge upgrade on the way that we process and the way that we run this through operations.

Our productivity is up and our cycle time is down. Overall this project on time. Um, quality out the quality project. Uh, and then we're seeing the people using it are getting the benefits. Our processes are just getting more efficient and timely. So in going on that, we went out and we surveyed some people within the organization who have different eyes on this project and how it was involved. And as we were doing this, we didn't know what to expect. You send the email out, you expect people to write back. But what I found really interesting, and I'm not going to read these verbatim to everybody here, but if you look at the people in our business areas, so like Tessa, um, Cathy, Deb, they're there or their teams are using this tool and they talk about how it's helped them and their efficiencies, how their teams adapted to it, how it's a huge improvement.

So when you see something like that, it really lets you know that you've gotten to the goal that we were trying at the very beginning. Then we look at Katie and Jen and Meg who were involved with the project, and you're hearing how the partnership between the organizations, how internally, how we ran through our SEO and discovery and UAT, the benefits that they saw, how the project really moved in the way that they were expecting. So getting all this feedback was really great to see. And then picking the ones that made the most sense to put on this slide. Um, it's the way we wanted to cap this off. And then I think I turn it over to dev. Yeah. Thank you Todd. So, um. I want to discuss for a few minutes about, you know, how are we empowering this Legacy transformation through AI?

How is Cognizant planning to implement such modernization with ease with solution accelerators? Right. And the common problem that we all have is Legacy platforms always teach us one thing to be more patient, more, more patient. And as Todd and Lynn spoke about, there are multiple systems that you have to handle a lot of documentation. And when I talk about documentation, the biggest problem that we all have is there are no documentation for many legacy platforms that have been built like decades and decades back. Forget about the things that were built like ten, 20 years back or 30 years back, mainframe application, or Lotus Notes. Even the ones that were recently built. No one has a proper documentation or process map, I hope. Like I'm seeing few nods here and there, so that is one of our biggest challenge, right? I started my career like writing thousands and thousands of lines of code.

And during those times we didn't sit and like document everything as like requirements or process map and things like that. Right. And that is a big pain point. Now when we embark on these transformation journey, right? So, um, from Cognizant standpoint, we really thought through this pain point. And when everybody has the speed to market as a key objective, like how do we, you know, like help our customers to analyze all these legacy platforms, which has absolutely no documentation at all, or even if there is documentation, we do not know whether it's up to date, current or whether the process maps everything. There is absolutely zero information in many engagements, right? So what we did was we created our team, created a Cognizant team, created a tool called orange. It's like the reverse engineering tool. It's one of the most powerful tools. It uses a lot of AI models, which basically extracts all the business logic rules and creates this beautiful document along with process maps as well, which can be fed to the Blueprint.

Right. I hope all of you might have played around with Blueprint. So our orange tool basically creates. That makes us easy and every step, right? The output of whatever tools that we create, we bring in human in the loop to validate, right? Most of the logic and the process are in, you know, like the business operations or businesses head, which has never been documented at all. So we bring in the human in the loop to validate all those documentation, the process maps, or the business logic that comes out of our reverse engineering tool called orange, and that is fed to the Blueprint what I think Blueprint is quite marvelous and all of you might have seen it. It automatically creates the skeleton of the whole application by creating case types, data models, properties, whatnot. Right. Um, we thought through like, okay, how do we build something that can supplement complement what Blueprint does, right?

Blueprint doesn't provide you the logic. And that was the, you know, like thought process for us to build something tool called build, forge, build forge has like two agents. One is called the build agent. Another one is a quality agent, right? When we are talking about a speed to execution, we need to think about like, how can I accelerate in every stage in the SDLC? Or these days, I call it as AI DLC, right? So the build agent under Build Forge tool, it creates more business logic. It creates sections, rules, your screens, and also business logic with source information. And it puts it in the branch so that developers, we are bringing the human in the loop. Once it creates that, like we are bringing a human to basically analyze, validate the rules, the sections and the data transforms that has been created from the tool, right?

And the beauty of it is it just doesn't throw. Okay, here are the list of rules. Here are the list of things. It also tells us, okay, here are the various case stages. Here are the various case type. For example, if it had created a section rule or any other rule, it clearly denotes more like traceability. Okay, for this stage. For this step in the process, here are the ten rules that I've created. Right? So that in the morning keynote, you heard a lot about like validation and governance. Having that level of information really helps us, helps any human to validate very quickly and also update more on top of what the tool had provided us. Right? So that is the build agent. That's all it does. And the next one is a quality agent. The quality agent basically creates automated test cases. And also it um, apart from creating the test cases, it also auto generates a test feed that could be fed into like automation, test automation tools like selenium and everything.

It's more compatible, right? And one thing I want to highlight about, um, both these tools are, they are model agnostic, meaning they, we have tested it with different models that are available in the industry. Like we have used Google and we have used Cloud. And they do work. And in fact, like we have given like how fast our, you know, like, um, delivery had taken, we had seen like 70% increase, right? Gone are those days when we present such huge numbers, the audience would even doubt or is it even achievable? Right? But in the world that we live, it's very easily achievable. And in fact, it is a more conservative number that we are highlighting as more models come in the market. Definitely there are more scope that it could be even done more, faster. Okay. I just want to highlight that these two tools demo are available.

The person who I dated who had built these are present in our booth. We are at Cognizant booth number 21. So please stop by and we'll be happy to, you know, like show you all the demo how both these tools work, both orange and, uh, build forge works. Okay. And, uh, with that, we'll open for the Q&A session. Um, I mean, please use the mic that's present on both the sides for any questions that you may have for any of us. Learnings that you had if you were to do this again. Is there anything you would do differently than what you started with in the beginning? The assumptions you had in the beginning, you ended out you ended up with what you wanted, but would you have done anything differently if you were starting over again on the next one? Yeah, sure. Um, I think, I think again, you know, one of the things I do want to point out, which, you know, led to kind of a pretty successful journey in this map was, um, kind of the, the Cognizant processes.

But I think, um, if anything, I, I really can't think of anything because number one, let me help you understand the context of our, uh, journey, I guess, um, from Lotus notes. Um, number one, it was under a severe time crunch. I will tell you that. And the one thing that you know, we would appreciate is that we were able to actually get a true reimagining, right? Because you can imagine under that time crunch that, you know, there's that, you know, potential or that that urge to kind of just kind of lift and shift and move. But I really do think that like, as part of our process, right. Looking at what part of the process was mapping out what's current and really examining, you know, okay, was this really necessary? And does this really make sense? And going to a true workflow that we really did achieve the best we could.

So at this point, it's really hard for me to actually, you know, think of anything that like, you know, through this particular journey that, you know, we could have changed. I mean, obviously there's always lessons learned, but I think this turned out really well with the process that we use. Yeah. Um, you mentioned, um, not doing a lift and shift. How do you assess from the legacy system, the antiquated processes that you don't want to carry over, the ones you have to reimagine? And if you use a reverse engineering tool, it's just going to tell you what you have. It's not going to tell you if they are correct good, wrong, bad, left, right and different. It's just going to tell you this is what what I found. So how do you, especially on a time crunch, how do you make sure that you're not just bringing old broken processes and forcing them into new technology?

That's a wonderful question. And that's why I said human in the loop is very important. We were kind of attacking the problem that to analyze, like, for example, if you take a COBOL code to analyze, like even a very small application, COBOL, like easily, it'll be like 3000 or 4000 lines of code, right? If whoever had worked in COBOL, apart from the regular templates, the actual core business logic itself is like thousands of lines of code. We were trying to solve that problem because somebody to manually go like a systems analyst or even a developer for that matter, to go and analyze those thousands and thousands of lines of code took months for few customers, right? And this gave you like within like few minutes, right? And that's a good starting point. Definitely one of our best practices, lessons learned when along with us, when we collaborated to reimagine the application, that's exactly what we would want to do here.

Orange to just extracts and gives you a here you go. This is a current state, right? And believe me, it. It won't be 100% accurate to. But at least even if it's 90%, it solves the problem of, you know, somebody analyzing the code for months. Rather, they have it within like a few minutes. And from there on, you have the foundation. From there, somebody has to validate along with business team and IT team. Then you have to come up with a two B structure, removing all the waste in the process and stuff, whatever that has been acquired all these years, and there'll be a lot in such applications. So that is how we are planning to do. And if I can jump in to. So on both questions. So one, something that go back on and I think Glen hit it when he was talking about, hey, we hit our objectives, but we were time crunched.

So that whole modernization, we did leave some stuff on the table because of the time. So doing it again, I'd like a little more time. So that's, that's what we missed. And then as far as making sure we didn't lift and shift, it's where when I talked about the discovery sessions up front at eight weeks where we did it, um, we brought thought leadership from around the organization. So it wasn't, hey, we're just going to talk to the enrollment people on enrollment. You'd have an appeals person or a customer care who will hold them accountable, say, I want to do this, this and this because I do A, B, and C, well, why would you do that? So by having that environment to challenge each other, help prevent some of the. The. This is how we always do it, so that a different set of eyes helped get us there.

And I think that is one of our best practices. And Excel has been very supportive of us because as you heard, you heard time crunch from both of them. But the amount that we took, the investment that we did in the front during discovery, I believe we strongly believe that's the key to our success and seamless execution later on. Yeah, we had a time crunch, but we exactly know what we are going to do up front, right from the apex of features and the complexity, what are the releases, how many MVPs? And we were very clear. So I think that is one of our key, uh, you know, best practices that we always carry for several engagements. Any other questions? Hello. Hi. Um, so I have a question about Pega capabilities. Um, so you said you had a limited time to launch. So within the time you had to transform, which Pega capability exceeded your expectations from your initial inception to when you started using the system?

And if you have to call out some capabilities that quite could not be there within the time frame, would you be able to tell us which capabilities really exceeded your expectations within the time frame? Yeah. Um, as, um, Len mentioned, compared to the Legacy and the modernized Platform, I think the biggest capability of Pega is like having everything giving a 360 degree view. So user need not like go into like multiple systems, right? Or even, um, doing a lot of things manually, right. So skill based routing where Users get the right task at the right time. You wouldn't believe like there have been times where just a work assignment every day for everyone without a proper case management platform like Pega, like the team leads have spent like 2 to 3 hours just to look at like, what is my work for today and who is going to do what?

And it may sound quite simple to all of you. Oh, okay. It took two hours. So what? Right. If the work doesn't get routed to the right person based on the priority or the skill, then there's a lot of a lot of impact for that, right? You may even pay a penalty because a case didn't resolve on time. Correct. So those are some of the key capabilities I would want to highlight. It may sound simple, but it's more powerful I would say. And of course Build for Change. Um, though we have requirements though, we have discovery during the course of execution, you will have some change. And how capable is a Platform. To adapt to change is one of the key factors that I want to highlight about going to the Pega Platform. And if I can, because from the technical perspective, one of the things I was most excited about was, right?

The integration ability. You heard me talk about API first strategy, right? So that was really key to like really integrating all these systems so that we could get to that single screen, right? So the extensive, you know, with the Kafka connectivity capabilities, um, obviously rest connectors, but you know, more surprisingly to myself, the API really allowed us to kind of extend like, you know, integration with each of the workflows so that not only, you know, I talked about our CRM, which is on Pega, um, integrating with those workflows, but also external systems, right? So for instance, you know, the facets of the world, right? So you know, where before, if you know, there's some teams that still kind of work working that system who have to now navigate. We now can even automate some of the workflows from external systems from fastest to automatically create cases.

And, you know, a lot. So that I think the API and basically a lot of the integration options really opened up the possibilities. Yeah, absolutely. Any other questions? Okay. Um. Okay. I want to sincerely thank Todd Len for your collaboration. And, uh, on behalf of Cognizant leadership, I'd like to extend my thanks for sharing this wonderful story. I wish you many more success and presentations like this as you embark on your AI journey to come. I wish you all the best and thank you. Thank you everyone.