PegaWorld 2025: Eliminating Friction: How Wipro Helped Lloyds Banking Group's General Insurance to Transform Claims Servicing

PegaWorld 2025: Eliminating Friction – How Wipro Helped Lloyds Banking Group

Today we are here to talk about our work together that we have done with Lloyds Bank, um, on improving transforming the general insurance claims servicing business. And I'm very proud and honored to have Phil Hirst who, who lives in UK with his wonderful family, has flown down all the way from UK to be on stage with us.

So proud and honored to have you here, Phil. And as for myself, I am Sanjeev Dubey. I lead the Pega ERP business for Wipro globally. It's part of our Digital Experience unit where we drive transformative experiences across total experiences, which includes customers, employees and partners. I'll let Phil introduce himself.

Hi. Thanks. I'm Phil Hirst. I've been with Lloyds Banking Group now for 27 years. So I started straight out of nursery. But in my career with Lloyds I've done many different roles. Being a tester, being a BA, being a project manager. Now the customer journey manager for the home home insurance claims journey.

Um, so I've had a wealth of experience whilst I've had those 27 years with Lloyds, so it's been really great career so far. Pleasure to have you here. Phil, thanks. Thanks for making the trip across the pond. Appreciate it. Um, a quick walkthrough of the agenda we'll walk through. I'm sure everybody knows Lloyds brand, the general insurance brand.

But Phil will take us through. You know, the details and depth of the coverages, the products and the markets that we cover. I'm hoping the same cases with Wipro we'll walk you through. At least enable yourselves what the work that we do and areas that we cover. Then we'll quickly jump upon to the challenges that this particular opportunity faced in terms of claims transformation for the general insurance.

What it presented naturally, the challenges presented, the opportunities, how we approached it. And we'll share between Finland, myself, Phil. All yours. Okay. So, um, for those of you who don't know, don't know, uh, Lloyds Banking Group is made up of several brands within the UK. There's 16 or so in total, but within general insurance we just operate under the four brands.

So the Halifax, Lloyds, MBNA and Bank of Scotland. So we've got a really rich history within Lloyds. We've been going around for about 325 years. So for those Americans in the room, it's about 70 years before your, um, declaration of Independence. Independence? So that's how long we've been going. Um, we've in the region of 27 million customers in total.

Most of those are mortgaged, so with the largest UK mortgage lender. So that gives us great leverage to get in there to, um, cross-sell their home insurance policies, of which we've got about 2 million at the moment. Um, that averages out about 117,000 claims per year. So we've got a lot of claims going through that are currently are fully manual process.

So, um. And there's quite a lot of cost to. Yeah, it is quite expensive in terms, um, of the running of the claims, we've got about um. 300 million pounds cost for our claims each year. So that includes all of the actual replacements and the running costs within the GI business. So just to go on a little bit more about the four brands that we operate under, we've got the three branch networks, so Halifax, Lloyds and Bank of Scotland.

So they're on the high street so we can operate through those channels. We've also do a direct to our own websites for all four of the brands, so people can buy their insurance directly with us. We also have the aggregator comparison websites where which we operate under. So we're on the go. Compare compare the market confused Moneysupermarket.

So customers have got a breadth of channels in which they can buy their policies from us. We do also, um, underwrite the insurance for some corporate partners. So the other brands within the UK that we work in partnership with. If you buy a policy for I, Sainsbury's, etcetera, that'll be underwritten by us and we'll manage those policies as well.

So that's a little bit more about just the general Insurance element within Lloyds Banking Group. We've got a really aggressive and ambitious, um, future ahead of us. We really want to become the number one in the UK for home insurance. We've set ourselves quite a high target. Um, and we want to be there not just for the sheer, um, market share, but in terms of the service, the policies, the efficiency that we can bring to it, to our customers.

So really driving to get that, um, built into our process. So we're looking at delivering, um, experiences with as much frictionless that we can for our customers. We want them to be able to when they need us most and making a claim they've had an incident in their home We don't want to add to that, um, difficulties that they're going through.

So if they come to us, we want to be able to handle their claim quickly, efficiently, with minimal force. Absolutely. And thanks a lot for sharing that. A little bit about Wipro. I'm sure all of you have already heard the depth and the breadth at which Lloyds Banking Group in particular, and General Insurance operates a little bit about Wipro.

Wipro has been 80 years founded eight years ago eight zero. We have presence in 65 countries, 230,000 employees, close to $11 billion in revenue. We pride ourselves in being at the top of the sustainability index for 16 years in a row. Extremely proud of our Gender Equality Index again continuously into the fifth consecutive year.

But the most important thing that we see ourselves is that 66% of our economic profits are irrevocably, irrevocably Are dedicated to charity. So our founder is, uh, he runs a charitable organization where 66, two thirds of our profits go. So we look ourselves as a socially conscious organization, developing the world together.

Very proud. Extremely, extremely proud of our relationship with Lloyds Banking Group as a broader organization. We've been working with Lloyds Banking Group for 18 plus years. 2500 employees of Wipro currently collaborate and partner with Lloyds Banking Group on a day to day basis. We have had relationships with Lbg across all the value streams, whether it is retail, corporate onboarding, security, lending.

So we have been partnering across all the value streams and continue to partner thanks to the support of Lloyd's leadership. Uh, looking at since this is PegaWorld, so focusing on the programs that Wipro has partnered with Lbg, the end to end simplification. This was the first groundbreaking Pega program at Lloyd's.

Almost a decade plus ago. So we were very proud of starting off with a Pega journey at Lloyd's. And then we have been engaged with the claims transformation that we are jointly sharing with you with ourselves. We are currently engaged in some of the economic crime prevention work that the team is doing, and term lending and overdraft. So we continue to support and partner with Lbg on these initiatives. So in terms of today's discussion, the opportunity that presented ourselves as Phil kind of alluded in certain areas, and he'll go in depth as we go along. The opportunity was all about a lot of these large organizations, and I'm sure a similar thing with Lbg and General Insurance Group as well.

A lot of it is on inability for the entire journey to be digital. Digital, when a lot of these organizations are starting, were limited to when you can just upload a document and then you cannot amend it, you cannot change it. If you have to in the back office work once. Even if it is digitally raised, it has to be manually keyed into the back office systems, which is SAP in this case.

The claim handling process is a lot in the people's head. So the fraud detection, the experiences that they determine is all based on what they have observed and seen. So a little bit of, uh, inconsistencies that could creep in. And especially if, uh, you know, folks retire or move out, those kind of situations.

So it presented an opportunity together to partner with, uh, Lloyds Group to figure out a solution that can be an omnichannel digital solution. We use Pega Customer Service for Insurance, uh, which which brings a lot of these capabilities from the from the grounds up and, uh, putting together an intelligent workflow that's going to orchestrate across all these systems.

I've got an architecture slide that I'm going to walk you through and identify the opportunities which which can severely accelerate the claims based on business rules and decisions that LBJ business leaders make. So we have embedded intelligence, artificial intelligence, and process intelligence into the process.

It's it's it's a journey that we have embarked. We are not we don't believe that we are at the end of the journey. We have started that journey together and embedding those intelligent engines, which I'll walk you through in a few fraud and anomaly detection, which can be done automatically. Going to Kara checking out.

So Sarah is the the State Regulatory Insurance Regulatory Authority checking out what the fraud databases etc. is out there. So kind of automatically putting that in the workflow and process so that people don't have to do it manually impacts at a very high level. There has been improvement in digital maturity.

People have started to use it a little bit more. They are able to do a lot more than just submit the initial claim. They are looking to amend it and the system can automatically. Do you know, the IDP, the collection, etc. so between Phil and myself, we'll walk you through into the details of what impacts have been created.

So where we were before we started this journey, we're using Pega. Like we said, we had a front end digital platform where customers could log a claim. It was a single journey. So regardless of the claim, whether it's just a drop TV or the houses had a major flood or the roofs blown off, the customer got the same set of questions.

Not great. And it was a very long process for the customers to complete. And then once it gets submitted, a colleague then had to rekey it into the system to manage the claim. Waste of effort. So, um, we wanted to get that simplified for the customers, for our colleagues. Um, we want to get away from that single set.

We want it to be tailored. So all of those claims, 100% had to be manually processed. So if we click on to the next. So where we are now we still have a standardized set of questions. But these are now processed straight through from the front end and into Pega which is our workflow. And that also feeds into SAP which is our system of record.

So that whole process now has been vastly improved. Colleagues are getting to grips with using Pega because this is a whole new journey for them. They've never had anything like this before in the claims business. So this is a real moment of truth where we're taking some of their autonomy away from them, which has been a real struggle.

They're now being told how the claim is going to get handled, what is next to be done on the claim, rather than them just working it through themselves? We always used to have a bit of a strapline that we don't have one decision engine. At the moment we've got 500, 500 decision engines all doing their own thing.

So the customers, you know, depending on who they get through to, they could be treated differently. And we want to make sure that everything is aligned. So regardless of who they speak to they're going to get the same excellent customer service. So that's where we are at the moment. We're getting, like we've said, some basic validation in there.

So as soon as the claims registered we can see whether they're actually valid. It's a claim or not. And we can either continue on with the claim or we can decline it straight away. We'll get some of the fraud triggers in there. We've got the documents being uploaded automatically now and before. That was a separate journey in itself.

But they can do that within the first time that they're notifying us of the loss, or they can do it afterwards. The customer has the choice. And then we'll do some immediate Validation checks on those using the metadata. So, you know, if the photo of their lost item was taken after it was lost. Well, that's a potential fraud flag straight away.

Because who can take a photo of a TV if it's already been stolen? You know, so we're getting those improvements in now in the future. Well, I mean, I've been blown away this week here. There's so many new toys that I want to take back to the UK and start implementing them into improving our journey even more.

So we're looking at all of the AI led validity and determination. What's the best route in. Does it go to a a personal, uh, field consultant? So somebody who's going to actually go out to the home, to the property to assess the damage. Can it just be straightforward cash settled without any other, um, checks being needed?

Or does it need to go to a a complex claim handler who can really support that customer when they really do need it the most. Like I say, we've got such a backlog now of improvements that we want to make to our claims journey. We want to be world class. We want to be the best. And I think with the support of Pega and with with with Pro that we've had so far, I think we're making the first steps towards that.

Absolutely. And so how do we approach and execute on a program of this size? If you look at the size and complexity of the organization, if you look at hundreds of decision engines, if you look at the nuances that are there in the products that exist. So we approached it a combination of starting with a study and research so that we don't end up re-implementing what is already running inefficient, which is the leaders are already seeing started with study and research. So our design and domain and consulting teams engaged from the very beginning did a full journey blueprinting which will have an opportunity to show you how it kind of evolved over a period of time. Decisions were made in terms of what kind of experiences will be shown in Pega, and what experiences will be shown as digital.

Uh, you know, kind of headless experiences to the end customers. The self-service channel will be headless in terms of react, etc. and for the employees experience, the Pega Customer Service for experience. Insurance was chosen as the continued platform of choices, right? Focus groups were ensured. They were created so that they can validate, as these designs are getting matured so that we don't end up in a direction which is completely off.

So as you would see in the timeline journey, uh, quite a bit of exercise in depth was done in partnership with Lbg and Phil's team to be able to achieve this. Right. Um, so this was all about as we were sharing earlier, this was all about end to end simplification, right? How do we look at the claims journey from start to finish?

How do we give opportunities to for them not only to attach the initial documentation, come back and reattach it, or if they don't have the document ready, can they have a link which they can just click and attach it? Right. So we kind of looked at it and at a conceptual level broke it out into front and back office, front office looking for reimagining the journeys, the visualization of those journeys or learning from what the the nuances of the products bring to the table.

Digital first notice of loss. What kind of questions you can ask. And the second bullet point that you will see over there is we have to make some architectural decisions. Right. Where do these questions sit? If you have to enable them and send them across different channels. Right. Who's the what's the source system for.

So those architectural decisions were made in consultation with Architectural Review Board of Lbg. Then Pega orchestrates the whole journey. It's not just about taking the claim and putting that into a queue where somebody is turning around and typing it back into the sap, which is the end system. Since we have enabled this documentation, upload, etc., you have to bring intelligent document processing.

It can make extractions of the information that is already embedded. It can pass that information to the workflow in the process. You go about the AI portion of it right. The process intelligence. So that's a journey which I've been sharing earlier. We started with a design collectively working with Lbg to ensure that we identify opportunities.

How do we accelerate? How do we make it snappy, how what kind of intelligence we can provide to the process? How much intelligence we can put in the process, and then the artificial part of it can evolve from plain old business rules to then it can evolve to machine learning. Then it finally evolve to generative dimensions of it.

So that's a journey we are on together. And in the end, um, you know, how do we detect frauds, you know, through calling. Right? Invocation at the right time, checking with the weather systems outside to see. Was there even a flood in that area? You are raising a claim claiming there's a flood which happened on a certain date.

If you are claiming it against a certain, um, you know, damage on the say it's a wind as your peril, then the idea is, was wind above a certain speed where it could be identified as a tornado or a severe storm, etc. so basically insuring because the coverages can be different based on your policy that you may have taken.

So insuring that this is not in people's minds or heads, it's all kind of in the system. This is the high level workflow orchestration. Um, and this this is really key. This this went through in-depth analysis. This is this something that we should be looking to do? Is this the right architecture or should we be looking for opportunities that some of the capabilities that are sitting in SAP.

Should they be brought over. So this went through enormous amount of iterations where we believe what's the right architecture, what's the right level of orchestration that should be in right. So as you see the bottom systems that support SAP's Duck Creek, Duck Creek remains the policy system administration system.

SAP is the system that is the the final system which is actually doing the payments or making the call outs to do the payments, etc.. So we didn't want to touch those. That would have been even a larger program. The intent was to see where do we bring in the efficiencies to achieve the end goals that Phil alluded upon?

Right. So, um, document extraction, Platform ensuring that you are able to make fraud detection call outs. You have a file system that's able to store the documents outside. You start with that journey of creating that Platform in Pega, which allows them to do the electronic, uh, you know, first notice of loss.

And it kind of orchestrates back and forth. So it's once a claim is created, it lets sap know that claim has been created or document has been uploaded. It goes to filenet. A link comes back to the system so it understands if it has to do extraction out of the document. It uses intelligent document processing.

And you see at different stages along the journey, there are intelligence decision engines that have been embedded, something that Phil alluded to. There are six of them right now, and I'm going to walk you through what they do and where their journey is in the current process. So, um, right now there are six.

The architecture is scalable. These six engines have been built in Pega. So what they are built modularized. So the expectation is that you can have you can extend them to more based on the need that is observed. You can tweak them. You can start with plain old business rules. You can configure them to have advanced machine learning intelligence based on how the prior claims were handled, what the final adjudication of those claims was.

So the first, most important part in the journey is ensuring that the claim is valid, right? You're comparing it against a whole host of scenarios. You know, you don't have pet insurance. You're claiming it as part of your first notice of loss. Those kind of basic checks, uh, system is trying to do that.

Um, does this claim even follow the the policy that you have taken against those, uh, you know, the selections that you made? Now, policy is coming from Duck Creek. A lot of those information already with SAP, but the intelligence that is coming in is marrying the answers that you have provided. You know, was there really a storm?

Did it happen? Was the flood against a certain level? I mean, you can't say if there's a flood, if it was, you know, half a foot of water and your house is already 5ft or 3ft above the base level. So doing basic checks. I'm just trying to ensure that or reassert that. Don't confuse it with the basic policy checks against the claim.

This is incremental to the fnol. Questions that are being put together. So system is checking against those. Then the assessment engines um there are required set of documents that are required against this particular claim or this situation. Calling out understanding what that list is. Ensuring that all the documents have been extracted, the extraction has happened, the comparison has been made.

Determining complexity. Because this is important for LBJ's business to determine should it go straight through. So that's the end goal, where the intent is that if it is determined to be simple and they can create other subcategories if required. But in current thinking, if the claim is simple, then can it be straight through processed?

Right. The idea is to see how how much can we slowly open up the pipe and responsibly open up the pipe? And then you have hybrid and complex claims, etc. routing engine, which is the right team? Um, if you already routed a claim to a particular claim handler and you see that somebody has removed an item from that claim, now the complexity changes.

It has become a simple claim now, right? But you would still like the claim handler to same claim handler to continue to use the claim. So all these pre-settlement options is it should I go to supplier one or should I go to supplier two? The same TV can be brought from three suppliers. What is the history of these suppliers with ourselves, with LVG, GenAI Insurance business.

So it applies all that intelligence comes back and decides what kind of settlement should be made. Should we be paying cash? Should we be asking them that we'll replace it with one of our this particular supplier that the pre engine has chosen. So these engines fire every time anything on the claim changes.

Somebody added a new item. Somebody withdrew the claim. Right. So these engines will fire constantly and guide the process to the most efficient execution. I'll let Phil walk through this user journey that was built together. Yeah. Thanks. So we've just simulated a journey here, um, from one of our customers.

Um, he's returned home, found a leak. He's registered his claim, so he immediately goes onto our app, registers the claim through the ethanol process, is able to upload some photos there and then because you can do that within the app. Um, and then what we will do then at that point it comes into Pega.

We'll do that. Like we said, the validation checks, do they have buildings cover? You know, um, is the policy still in force? Is it paid up to date? All of those kinds of checks that immediately we can do to to automate the process. Once we've done that, it goes into a Q, and a colleague will then pick this claim up.

So we've bypassed the simple colleagues, as we call them. Um, but they're the ones who are, uh, new into the business. They, um, are less skilled, less experienced. Um, so that's a simple claim. We are trying to get away from calling them simple claim handlers because they're not simple. They do go through an awful lot of training, but it's a simple claim, i.e.

low value single item, that kind of thing. But in this case, because it is a flood, it's gone on to one of our hybrid claim handlers. So they're a little bit more experienced there. They've got higher authority levels to make settlements for the claim. Um, and they've assessed it. They've got the evidence.

They've reviewed it manually as well. They've checked the metadata that's been extracted, checks that everything's okay, and they've got to a conclusion where they're happy to settle this claim. So there's one of two choices here. So depending on what the item is, the level of damage, it could be that we just agree to get for the customer to get an estimate.

And then we can settle at that and we'll, we'll check that estimate against our schedule of work. So we know that we're not being, um, overpaying the customer. Or we can get one of our suppliers to go out and fix it for them. So in this instance where we have that conversation with the customer. We agree we'll get one of our suppliers out.

So a plumber goes out and, you know, makes good repairs to damage for him. As far as the customer is concerned, then that's it. It's done. You know, um, he'll he'll have paid his excess that he has to pay, but nothing else. The supplier will then invoice us directly, uh, for the cost of the claim. We will settle that.

That's all been done between SAP and Pega as well. So we're automating as much as that of that process that we can. And then obviously customers happy we've settled the claim. We've tried to do it in a shorter period of time as possible. This is very much dependent upon how quickly the customer gets their, um, estimates or documentation to us, but we aim to get these claims in and out, get the customer back to where they were as just as quickly as possible.

Um, so yeah, that's a very simple abridged version of a claim. Obviously we've got other variants of claims, where if it was an accidental damage to a TV, we could potentially from the customer register and the claim uploading the document settle that claim with them within a matter of minutes. You know, we're looking to really streamline the process.

Um, and so that the customers inconvenience the least possible. Okay. So this has been quite a long journey. You know, we started back in 2022. You know, we have gone through several iterations of, um, rolling out change into the, um, claims business. I think the key thing that's taken us, and I'm going to preempt a question now as to why it's taken us so long.

Um, I think we kind of underestimated just how complex our own products are that we sell. And that has been a bit of an eye opener for us. Um, so, you know, that's taken several iterations to get us to a place where we're confident with what we have have delivered. The last thing that we want to do is decline a customer's claim when they're covered.

That's the worst possible outcome, not only just for the customers, but for us, because that would be a definite failure on our part. So, um, we did the initial rollout of changing the ethanol process to using Pega back in November 2023. Um, last year we, um, expanded the end to end processing to. So colleagues were using Pega then for the full life cycle of the claim to process it, because what they did initially was keyed it into Pega, but then managed the rest of it after ethanol still in SAP.

Um, so we've got the end to end processing. Just recently we're expanding it out because if you remember back, we have four brands. We did it on the Bank of Scotland brand. Just that's our smallest book size. So it's a bit of a safe environment. You know we don't get too many claims coming through there.

So we could really babysit each one of those claims fully, end to end, just to make sure that everything was working exactly as we expected it to. Um, like I said, we're expanding to launch on all of the brands imminently. And then from July 20th this year onwards, we are starting to do some straight through processing.

So, um, low value but high volume claims. We're going to just get the whole Pega to do the whole process. So claim will go from start to end without even being handled by claim colleague. So we'll start now on freezer full claims. So if you if you freeze Freeze is damaged and you've lost all of your steaks, your lasagnas, and everything that you've got stored in there.

Uploading a photo. We'll get that settled straight away. You know, like I said before, that claim could be settled within minutes. OK. Which is a real advancement on us. We'll then once we've got the confidence there, we'll roll it out to other perils, other types of claims, other higher value items as well.

So we're really going to start pushing the boundaries on that one. So yeah key learnings. We were ambitious to start with. We wanted everything and we wanted it now. And we just realized that there's just too much for us to go at. We needed to, um, stop, think about it. Plan it out better. Small incremental improvements is better than trying to do a big bang because we just needed to find our feet.

You know, it's the first time within GI that we've used Pega, so it's a massive, steep learning curve for us. Um, the ops, the operation team. Um, we should have got them on board earlier, because not only to get them used to the system, but to get that mind shift because, like I say, the system is going to do a lot of the thinking now.

They're they're losing a little bit of the control that they have over how the claims are managed. And we needed to get them on board a little bit sooner so that they can feel confident that the system is doing what it's going to. It needs to do for the customers. Um, covered off the release planning we wanted.

I everybody wants I, we wanted it three years ago. We just weren't ready. Um, I think now we're getting into a position where we get more data. We understand what claims are coming through now because we're getting so much more granular details about our claims, whereas before it was quite a high level, the colleagues didn't put every single item on, whereas now they do.

So we got that level of detail that we were lacking previously. So we can really use that to drive through our settlement options and what's best for the claims. Yeah. And to add to that, um, you know, one of the key things, once you have the quality of data in the claims, uh, you know, uplifted or new incremental metadata is captured naturally, it's going to lead to the machine learning elements of it, how you can ensure that you can build machine learning models that will be able to add, augment to the plain old business rules or processing in the different engines, etc..

And later on, it's going to come back to the GenAI portions of that. So a couple of key learnings from our delivery perspective. Uh, naturally, Phil alluded upon, uh, from the product and the size and the depth and complexity, etc., from a delivery of program. As you can see, enablement of the product features.

I think there was a appreciation that came along over as the project progressed, that the complexity and the nuances and the variation, including the people like Phil, who has spent almost three decades at Lbg, there was a lot of learning that came along. Would you like to touch upon Phil, how it kind of evolved? Yeah, I. Mean. I mean, the way our guys have been great. You know, they're the experts. We've had to put our trust in you guys to deliver us. It's just such a massive shift for us to come onto a system like this. Including the complexities that you saw in your own products. I mean, yeah, our products are very, very complex.

So just getting that actually for ourselves, it's a bit of a wake up call, to be honest, for us to say, just see how complex our product is and if it's complex, complicated for us, it's going to be complicated for our customers to understand as well. So that's something that we need to take away and see how we can look to improve that.

Absolutely. And then I'm sure folks who have been doing delivery for a while recognize that integration is the longest pole in the tent, in this case sap. Duck Creek is it took some time. I'm not saying it was difficult or hard, it's just a matter of planning. Better planning could have worked well.

Project team. This is a our share of experiences that we are eager to share. Enabling the product team, project team on Pega capabilities so that when they're writing requirements and evolving those requirements, they can be more aligned to what is out of the box so that somebody is not asking you to create audit trails and things of that nature.

With Pega automatically does that. You can significantly shrink your discovery and your timelines I models. I think when the word AI gets spoken, the expectation is it's kind of a turnkey thing. It's a journey and that's something that we have been on to it together with our clients and Lbgq Insurance.

Right. So they'll yield results. Definitely yield results. They're dependent on the quality of data that is available and that will drive the results. So I'll let Phil cover the results that they have been experiencing in the rollouts that have already happened. And before Phil, you go there. But I'm extremely thankful to him that there's actually a rollout that is happening this this weekend.

And we had to literally pull him out to share the story here. So we truly appreciate it. Thanks. Um, yeah. I mean, the changes we're starting to reap some of the benefits of the past three years. Um, the simple fact that we're now down to one decisioning engine, which is Pega, rather than the 4500 colleagues that we had.

That is a groundbreaking for us, you know, to have that level of control over the claims. We're getting faster responses. You know, the claims have been registered far quicker now, simpler for colleagues. The key in a claim into the old SAP system. It took them forever to do that, whereas now it's so much more intuitive using the SAP Pega system, and it still creates the claiming SAP, still creates all those sub claims within there, but they don't see any of that now and it's great.

Um, getting lists of what the actual correct settlement approach is for each individual item on a claim. Again, groundbreaking. It was down to the colleagues best, um, understanding what they what what they thought was best for that particular claim. Whereas now we've got the strict rules behind it.

Um, the amount of digital claims that we're getting now is going up. You know, we we're at 70% of, uh, increase now over what we were a couple of years ago. And that's because we've opened it up to all of the channels so they can register a claim through their internet banking. They can register it direct on the public website when they log in. So getting that opening up those channels Nils is actually freeing up our colleagues to deal with the more complicated claims. They're not dealing with the mundane as much, and it's more rewarding for them. They can get stuck into a really complicated claim rather than doing lots and lots of manual, tedious tasks.

So it's a lot better for our colleagues. They they enjoy it a lot more. Um. All of the these just come about and really vastly improve the customer and colleague experience. Because one thing that I've been pressing on is we're making a change to for colleagues as well as our customers, they're reaping the benefits of using Pega.

Now. It's simpler, it's more intuitive. It's it's a better working day for them rather than being directly into, um, sap. So probably we are at the top of the hour. Any questions? We appreciate. If there are additional details that you may like, then feel free to reach out to folks at Pega or ourselves at Wipro.

We'll be more than happy to share, but if there's time, then we could take some questions. Any questions? Yes, sir. You might need to come to the microphone. Okay. Do you use Blueprint at any time during the the process? So that's a great question. And that's we are being asked by a lot of our other customers as well.

This journey started in 2022 as you saw. So it's kind of evolved. There are opportunities if the extensions that come along to newer journeys. But the short answer for this particular journey, it wasn't it started before Blueprint evolved. And for the for next steps. We think the intent is to use Blueprint.

Yes. Thank you. Okay. I think there's one more gentleman. Gentlemen. Quick question. Um, regarding the, uh, the the omnichannel experience for the client coming via the self-service channel versus the internal staff, the employees of the client coming via the application portal. Um, I heard you said it's a react based portal on the self-service.

For the self-service piece. That's headless. Headless Pega invocation. That's correct. So, uh, why didn't you consider lift and shift of your, um, workflow with them, but web embedded mashup? So that's a great question that was evaluated as part of our architectural decisions. The reason they had is they already had a strategy for the front end channels, which was pre-decided as an architectural decision across the board that all the applications that will be invoked through for, for now and for future as well, those invocations will always happen, uh, headless into those systems, regardless of Pega or any other experience that they were building and those experiences, those websites and those kind of things already preexist in a certain architectural forum.

So the team did bring up to the attention to evaluate, and that was discussed the value it could bring. We naturally believe that it could have accelerated few things, but we also wanted to ensure that we stay with the futuristic architectural pattern. And the other question I have is chatbot. So how are you integrating the chatbot with the LLM?

Is it going through from the self-service channel perspective? When client comes in with with a claim request creation, they see the chatbot how the request and response flows through. So did you mention LLM or did you? I may have missed the chatbot. You. Have you integrated with GenAI, right? LLM. So not in this yet. In this. Yeah. So the AI and the AI machine learning elements in the decision engine are in the upcoming phases. Right now. Our interest with the group's organization was to ensure that we roll out to as many countries and products that are out there. So the journey was really if you saw that slide, it was initially the Fnol first.

That was where the bigger problem was in terms of customer facing inconveniences, that once that was addressed, then it was end to end. And once end to end is addressed, it's primarily feeding, not GenAI yet into the field. But the way it is architected is it will always extend into the ML machine learning engineer interventions.

It's already thought through that's coming, but it's not ready yet. And how is the performance of your self-service? How many APIs? Once you'd asked that question to me in the lobby, I'd asked, how is he feeling? So he's the same performance question. Do you feel anything about? No. I mean. Performance.

System performance has been great. It's fine. We're not seeing any issues at the moment with performance. So it's not Pega Cloud two. So we're leveraging Pega Cloud. Okay. Thanks. Thank you sir. Okay. I have a question. Uh, are you using a Process Mining or BI tool to measure the process, to understand?

You know, you have long way to go, right? Yeah. So are you measuring at all, like, where are current bottlenecks? Uh, to understand, you know, what needs to be improved in the future. So if yes. Then. How does it connect to Pega? So we're doing extracts and we're getting data out and putting it into power BI.

So the business teams are really carefully closely monitoring the whole process at the moment. One of the key things that we do want to get out is for each task. How long has it taken compared to how long it used to take. And then we can start to use that in our forecasting models to see actually what is the mix of the claim handlers that we've got.

We've got the main at the right time. Obviously we have spike in claims during the winter months when we get more storms. So we want to make sure that we got the right people in the right place and that the claims have been handled quickly and efficiently. So when things like a storm hits, we can tweak the settlement approaches so that the, um, lower colleagues have more authority in which the payments to take, so we can tweak how we handle the claims at real time.

And what we also do is give customers an automated go ahead limit. So if they have damage to the roof due to a storm, we'll automatically send them an email saying, right, go ahead up to a certain amount. That way they're protecting their property. It's not being made worse because that's ultimately going to cost us more money.

So they can go ahead, get a local repair person in, fix the property, make it good, make it watertight. Therefore, uh, preventing further damage. Thanks. Thanks a lot. Thanks, all of you, for attending. Thank you. Thank you. Thank you.