Overcoming legacy debt and activating innovation with Gen AI, An Interview with Denise Garth, Chief Strategy Officer, Majesco The Leadership in Insurance Podcast ยท 26 Aug 2024 Source: https://shows.acast.com/theleadershipininsurancepodcast/episodes/overcoming-legacy-debt-and-activating-innovation-with-gen-ai Transcript: https://podcasts.opennash.com/shows/leadership-in-insurance/overcoming-legacy-debt-and-activating-innovation-with-gen-ai-an-interview-with-d-516e31/ (via OpenNash Podcasts, machine-generated) This is the Leadership in Insurance Podcast, brought to you by Finpro Search Partners. Insurance companies are businesses and they need to look for the long term and be sustainable. We went from zero to one and now it's going from one to a hundred. Insurance as a concept, as a kind of service, is brilliant. The execution is what we're looking at now. I think the companies that are going to succeed are the ones that are going to understand and master the art of intent. When we talk about innovation, we lean too heavily to think about technology and we don't think about creating a culture of innovation. I think innovation is essentially continuous improvement of existing processes and platforms and product, right? It's got to be easy. It's got to be seamless. Hello and welcome to the Leadership and Insurance podcast. podcast um this is your host alex bond and i'm very lucky today to be joined by denise garth who is chief strategy officer like majesco um denise thanks for joining me how are you i'm great how are you i'm all right although as you can sell i'm a bit i'm a bit shiny and we're just sort of doing that bit where we do when we're british in the summer we moan about how hot it is um and then we complain that it's not hot the rest of the time so it's one of those particularly hot days today so other than that i'm very good i'm very good um i've wanted to do this for a long time uh because if i started reading out your accolades as like 50 most influential people in insurtech uh but top 100 people that listen to an insurance or i think we'd be here for half the recording so um yeah i'm really grateful that you've had to some time to find uh with us on the podcast today but for the people that have been under a rock forever um as is customary on the podcast. It'd be lovely if you could introduce yourself and obviously Majesco for anyone that doesn't possibly know who Majesco are, which I think will be a small part of our audience, but you never know. So, you know, it's Denise Garth. I'm the chief strategy officer at Majesco, also responsible for our marketing at Majesco. And I've been here for over nine years now. And it's been really part of a transformation. My background is, is I was in, I've been in insurance my entire career. I was inside insurance companies, both on the business and the IT side, the first 20 plus years. And then I moved over to Accord for about eight, nine years there. I went to another technology vendor that also did BPO. And then I became an industry analyst with Strategy Meets Action for a couple of years. And then somebody who had hired me into a previous job, called me up and said, Hey, Denise, do you want to have lunch? And my husband said, Oh, you know what? This is about Denise. And I said, yeah, I do. So he pulled me into, into Majesco. And we had just done one of our first acquisitions of cover all at the time. And over that course of this period of time, we've done a number of acquisitions. We've been bought by Toma Bravo, one of the premier private equity firms. It'll be coming up on four years. and we have completely transformed our solutions in addition to the acquisition of a number of our other solutions. So it's really a really strong portfolio of what I like to say our next generation business solutions for insurance that really, first of all, are native cloud, but have a lot of other attributes that I'm sure we'll get into. So that's kind of the background for Majesco. We serve both the property and casualty and the life and annuity market. We have core solutions for both. We have loss control, digital solutions, distribution management, underwriting platform for group and worksite and benefits. So it's a wide range of solutions that, quite frankly, really does follow the end-to-end value chain. And it really gives us an opportunity to work with customers depending on where they want to start their transformation or where their pain points are that they really want to do something. And then really develop and build off of that relationship. Amazing stuff. Yeah, that lunch conversation. I love those. That's our biggest competitor. Someone getting pulled to lunch by someone they used to work with in the search industry. Yeah, it is funny. Someone said to me the other day, what's your biggest competition? I was like, the networks of individuals. I mean, always. And, you know, we're adjacent to that. I, yeah, I wanted to start with core systems because, you know, it's no secret that we have a pre-call to talk about it. And I was really intrigued by the way that you really broke down how you think about core systems, you know, Majesco. So what makes up the dimensions of the products for you today? Because I just love the way you articulated so kind of clearly and so distinctly when we spoke. Not putting you under pressure because you've got to do it again now. No, I've got to do it again now. I think, you know, when I think about core systems, you know, you think about it in terms of policy billing and claims. That's really what you've got. But we think about it much, much broader than that. So maybe if I could kind of back up, because I think to put in context of what I think core systems are today, you need to understand why they have to be different. So I think from an insurance industry perspective, we're at an extraordinary crossroads. We've got all kinds of pressures. There's a lot happening. If you want to talk about climate change, you want to talk about increased risk. You want to talk about a different changing set of demographics of buyers. There's all kinds of pressures that are really, I think, signals for insurers to say, step back. We got to rethink your business operating model and your technology foundation to remain relevant and leaders in the market because everything around you is changing. And that in itself should suggest that your business model and your technology foundation has to change. So with regard to technology, you know, we have accumulated what I like to call legacy debt over decades. And it's not uncommon to have systems out there for 20, 30 years. And that's an off balance sheet set of technology work that has to be addressed because we haven't really managed it well. We have just kind of put pieces onto it to try to make it as effective as possible. But ultimately, it's hamstringing insurers' ability to compete. So what we're seeing now with all the really challenging financial and market dynamics is that profitability and operational costs remain the top two mind of issues for insurers. and to compete not just today but in the future expecting anything any value to come out of legacy is really going to be is is really kind of a pipe dream because if you're stuck on that legacy and I call even on-prem modern systems we you know in the last five to ten years everybody was going through a transformation project they implemented a new core system and it now is legacy because number one, it's on premise. It's not in the cloud. Number two, they highly customize it. So it's very difficult to upgrade. And now what has happening is that insurers are not boxed in. It's limiting their potential. And so I think for leaders, market leaders to really seek operational effectiveness options and to capture, you know, the intellectual knowledge and the information that is part of the organization, it's unique to the organization, they are going to have another challenge and it's the retiring of employees. We think about core systems have to be what I call next gen and intelligent. And it has to be really be built on a different type of architecture. So the key components include a cloud native architecture that you really leverage that full potential of cloud computing to make it scalable, to make it containerized so that you can really have elasticity and automation that allows you to kind of shift as the demands in the marketplace shift. It's got to be open API standards. That means it can seamlessly integrate with your internal systems with a robust library of APIs, but more importantly, you can actually integrate with external systems. So other insure techs out in the marketplace that have really unique capabilities that you want to be able to leverage to create different kinds of experiences or to do some type of data and analytics. It's got to be headless architecture, meaning that it allows you to have flexibility and adaptability to stay ahead of the competition. And then you've got microservices. That's really important because everything can kind of come down to microservices, and it really allows you to have kind of a DevOps and an infrastructure and a set of security that can really be consistent and be managed much more cost effectively. But the most important piece that really, I think, differentiates this next generation of core systems is that data and analytics are not an afterthought. They're not a bolt. They're not a, oh, let's do this now. No, get about it as a foundation of the core. And so what we've done is we've embedded the data and analytics into the core. What does that mean? That means that now for the first time, our customers who have our core solutions have access to all of their data from that core system in real time. It's put into a data lighthouse. You can add other data to that. You can do other things with that. Secondly, you've got business intelligence. Remember, business intelligence was the hot thing the last 10 years. You have these BI reports that can give you a dashboard to tell you, you know, what's your business looking like? How many claims do you have? all of that kind of stuff. That was kind of like an add on to nope, we've embedded it into the workflow, we've embedded it into the core so that not only do you get the out of the box dashboards, you can actually create your own dashboards if given the authority to do so. So what does that mean? If I'm a claims adjuster for a certain line of business or for a certain region, I may have some really unique stuff that I want to see a different type of dashboard than my peers do so that I can manage that portfolio business in a very different way. Then you've got the AI machine learning models. There's some great partners out there that we've got partners that we're integrating and embedding those within our core, but we have built our own models as well. And that needs to be a part of the business process. So whether it's underwriting or it's claims or it's billing, you can leverage those models to be able to drive better decisions and to have better outcomes, business outcomes. But the most important piece and all of those other components of the architecture are really based on, have to be in place to be able to leverage this is Gen AI. What Gen AI can do is it can really allow you to drive what we have not seen since we first started automating back in the 70s and 80s is a level of productivity and a level of operational optimization and efficiency that we have not seen for decades. So what do I mean by that? If you've got Gen AI, the way we've used it is we've got Gen AI embedded. It's on every screen of every solution. And what you can do is that I'm a new employee. How do I do an amendment? How do I do a quote? I can, all of our online documentation is in there. We've run it through one of those large models and it can come back and it give you step one, two, three, four. And so for a new employee, all of a sudden your productivity is going to increase because you've got an assistant guiding you through the process of how to use the system. Second of all, you've got the ability to be able to use Gen AI to do some tasks and activities. Can you summarize my claims notes? If you went in and tried to look at all the claims notes and try to manually summarize that, it could take you 20, 30 minutes, maybe more. It could be done in less than 60 seconds. Those types of things are going to drive a 10 to 20 times productivity level that once you begin to build those capabilities in, all of a sudden you have a different operating model. You then all of a sudden have a different operational cost structure. And all of a sudden you have a different kind of competition from a product pricing perspective that positions you in the marketplace more competitively. And in my long career, being in the insurance industry the entire career, I've never seen this level of opportunity ever within our industry to have this level of real impact to the business and ultimately to our customers out in the marketplace. It's so exciting. I mean, just jumping in there with the, you know, not all carriers are created equal, but, and obviously people want different things from their systems. But, you know, one of the challenges of this conversation is always going to be the breadth of things that Majesco can do. But what are the things that they're most interested right now? What's the, you know, when you're having those carrier conversations, what seems to be front of mind for, you know, thematically of all those things that we've kind of broken down and covered off? What are they focusing on right now in 2024? I think everybody's focused in on their operational costs and their profitability. I mean, particularly on the PNC side. If you're on the life annuity and health side, it's a different because high interest rates are driving increased interest in products. Although on the group and benefit side, it's a little bit more challenging because people don't have as much income. So there may be, you know, selecting fewer benefits. So the financial and market dynamics have a different kind of aspect. So I think when you look at it, it's different for different organizations. And, you know, one of the things is, is that PNC insurers are trying to cut costs. We've seen some layoffs in the marketplace, kind of unheard of in some cases. You've seen a change in cutting of budgets. You can't cut your way to, to growth. You can't cut your way to, you know, you can cut to profitability, but you're going to cut off your, your nose to spite your face. at some point you're going to have to start investing this we've waited too long and now we've got this um accumulation of debt that now is really kind of having a big impact on our bottom line i think alex if you kind of go to um am best innovation scores and they had a report come out in may and they actually for pnc they actually highlighted where those that had higher scores of innovation, there are actual financial benefits to having invested in new technology and innovation to the tune of better operating costs. You've got decreased combined ratios. I mean, you have better growth. There's all kinds of real financial data now that's kind of starting to come out that shows that that investment does have a value and does pay off. We just haven't made those decisions to do so. And I think that depending on what lines of business, you know, insurers have, their priorities kind of change and where they're at in their own transformation journey. But I can tell you, we still have a lot of legacy debt out there, even the big team, because they got a different system for every line of business. Well, I was going to say that it's something we've talked about before, but life insurers, are they much more challenged by legacy tech than just by the nature of their businesses? Is it a bigger problem there or is that something PNC likes to tell itself? I think it depends upon the organization. I think PNC can still be running old, you know, 15, 20 year old systems too, particularly some smaller insurers. But life insurers tend to have them for 30, 40 years because it's hard to take the business off the books when you've got a whole life policy or universal life policy sitting out there and you're waiting for the person to die 50 years later. um you know and and typically once again those systems were very very um customized and um you know lots of customization done to them that you can't really upgrade them and it's hard to convert them the the data question you know we're talking about gen i gen ai we're talking about kind of analysis of data um you know the state of play generally like and they even access the data at moment you know like how bad is it for some of these people because i remember it wasn't even that long ago me doing a horrifying search in the claim space and i put someone into a very large reinsurance business and they were sort of saying where do they get their updated claims data and and uh i placed this person who's head of europe um and he said he phoned me up and said alex you would not believe where we get our claims data from and at the time they were getting it from brochure because basically the system wasn't capable of filling it out efficiently and now that was 10 years ago so i suspect that's very much an outdated thing but you know can people access the data in the way they want and and then i suppose where i'm going with the conversation is is that how big is the challenge is it getting up to speed because it's all right to get new tech and go forward and say well we can collect data from this point onwards but can they even access the data they need to kind of make that transition effective? There's a real problem with data and having access to the data. But even the bigger issue is the quality of the data, is the data of inequality within it. So if you're talking about legacy systems that are 10, 15 years or more older, more than likely, you have limited access to the data. and you're trying to write all kinds of extracts to get that data out into a data warehouse, into a data lake. But sometimes you may not be able to get all of that data. Even some of the modern systems, I would say up to, you know, three, four or five years ago, you only got access to the data that the vendor allowed you to have access to, you know, based on whatever they built an extract for or based on whatever their data architecture was. Today, I think, you know, if you are really building out a new solution into that next gen architecture, you should have access to all your data. So if you're looking at a solution and you can't have access to all of your data, that's your most important asset as a company. And you need to really kind of reconsider is that solution and it's bells and whistles that much more important than having access to your data. And you've got to have somebody who, you know, if you've got quality of data and somebody who's really responsible for that, and you can really make sure that the system that's producing that data is accurate, that's even more important because bad data in is bad data out. And I think that's been one of our biggest challenges as an industry. Could you really rely upon some of the data from a quality standpoint. What do you think all these data demands mean for the industry from a talent perspective? You know, do we need, obviously from a tech perspective, we need different skill sets. But even just kind of looking at the underwriters, the claims professionals, or is it less that we need different skills, but the training has to change, you know, the emphasis on data, the importance of data and the care, therefore, of being like custodians and good custodians of that data clearly has to change. So I don't know whether it's a skill or an emphasis of training or maybe it's a bit of both. But I kind of wanted to get your thoughts on how this changes the roles of people within insurance. I think it changes it a lot. I think that, you know, we as you move into solutions that can really automate far more using, say, a gen AI. to do things much more efficiently and effectively. And you have the technology that can really automate it, what you want, or you don't want people who can just do, you know, run the process and manually do it. And then maybe you do something off the side and you kind of come back in, you data key it in. You want somebody who are knowledge workers that are going to be able to make decisions in the process. So if I'm a claims adjuster and I'm looking at the information, How do I get that summary? How do I bring up a dashboard to kind of assess it? How do I run a model that can assess, you know, the adjudication of stuff, litigation possibilities, subrogation possibilities? How do I kind of manage that portfolio business differently? And so I think what all of these technologies, in particular AI and Gen AI, are going to do, it's going to give people on the front lines the information they need to be able to do their work much more efficiently, but more importantly, much more intelligently. They can make better business decisions on behalf of the company and on behalf of the customer. And so I have no fear about this. I know that there's a little bit of fear out there. It's going to take away the jobs. You know, that's kind of, you know, some of the conversation out there. I don't think that's the case at all. I think it's going to just provide more insights for the knowledge worker. It's going to make it more valuable. And quite honestly, when you look at this next generation, they want a job that's going to be something that's of value, something that's meaningful. We have one of the most important industries that's really meaningful. It puts people's lives back together. It puts people's businesses back together. It helps to protect. It's really the underpinning of lives and of businesses and of economies. It's really an industry that's really, really important. What better opportunity for this younger generation to be part of something that is really important to the livelihood of businesses and people. I agree. You're preaching to the choir here. I think it's really funny when, you know, we, you know, we were a business that specializes in insurance and insurtech space. And, you know, we've actually onboarded a couple of new people this week. And it's quite interesting seeing the cogs turn when they're in it a bit longer. And then, you know, Callum is a, one of my colleagues been with us six months now. And now he's kind of coming back going I saw this thing on the news and I realized that the impact it would be and then he was like and then he was asking me about we know what would you think the impact on reassurance would be and and it was really interesting to see how kind of interweaved it was with society and and and that's with that's looking at things just from a kind of claims and interest perspective but then when we started digging into kind of yeah bridging the gap in kind of coverage and uh you know and that and and all those good things and we we probably don't do a good enough job at selling that externally so I do think that's interesting I mean and the other thing that I think is interesting that we don't do well enough is that you know you've touched on it here we do a bad job at talking about the exciting things we're doing with technology so you know AI is the buzzword at the moment there's all sorts of different applications but actually real world applications really kind of meaningful applications I wouldn't say they're few and far between but what I mean is I think some of the a lot of the hype is about stuff that's kind of doesn't matter but what we're saying is that insurance does matter and then the application of gen i gen ai and there is really important what i wanted to kind of ask you from a majesco perspective why is it important that gen i is embedded why can't i get a bolt on off-the-shelf solution that offers something else um yeah why does it have to be part of the suite of my existing products it makes it But it's Gen AI. There's a difference between AI models and Gen AI. AI models can be a bolt-on because you're going to do, you're going to look at data for a subrogation element. That's one little part of an overall claims process. You can kind of integrate that in. Gen AI is really an end-to-end process. It's not just about a point in time, because if you're going to be adjudicating a claim, there's lots of different processes that you're going to go through to be able to manage that claim. It's not going to happen in a once and done type of thing. It's going to happen over a period of time. So you've got to kind of think about it on an end to end basis and you've got to have the consistency of the information and the data that you're using. It's embedded within that core system that you're actually running those natural language models against. and so I think that that's the big difference I think both are needed but they're needed in different ways so I don't see Gen AI at this point as really a bolt-on because if you can say I'm going to do we actually had a customer at our product council and they were seeing live demos of our intelligent core and they have policy and billing but they don't have our claims and they were so excited. They said, oh, this is going to be so exciting. And I said, yeah, what are you going to do when all of a sudden you get from policy to billing and you've got all that intelligence and then you get over to claims and you don't? And he goes, oh, crap. You know, because all of a sudden it changes, you know, how you're going to train a whole different set of people within the services department and the different solutions that they've got. And I think it changes the conversation as to, well, I want to be in the policy department because that's going to be a lot more easy to work with versus maybe the claims department it's just it's going to be different yeah yeah no I mean I think that's why I was teeing it up because the perception is that you could just plug these things and and it's like no it's not it's not end to end so like it needs to talk to each other it needs to kind of have this cohesive overview that goes between leads yeah it needs to be trained I mean and I think that's that's the challenge for leaders at moment as well is you know you're a leader in an industry that you understood completely and now you have to be a leader of a technology business at the same time and you need to understand it well enough to go what am i what are my solution options and and how can i apply this in a way that's meaningful um and and efficient um because otherwise it's really easy to i mean the amount of ai driven solutions i'm being offered at the most in my industry it's hilarious um i know almost all of them are bad i mean uh you know and i love trying these things and you know i've been involved in some of the R &D processes and being part of product councils myself. And, you know, it's what's interesting there is it's quite nascent. So there's a sort of lack of understanding of like even what we do for, you know, we kind of point things out and go, that's not actually how this process works at all. So we're going through the growing pains there. Just to add to what you said, I think, you know, because there's such a rapid adoption and experimentation of AI, machine learning and gen AI, I think what's happened is that data and analytics is no longer just a long-term strategy of incremental investment. It's a near-term reality that really requires a much more broad, holistic approach to data and analytics so that if we are going to add things on over time, we have it based on an overall strategy and not just, oh, let's try this. Oh, let's try this. And it's just a patchwork of stuff rather than an overall foundation that's built upon that's going to be really solid. You know what I mean? Yeah, quite. I mean, I think that's where we were coming from it. Because, you know, thinking about our tech stack, you know, people think, oh, it's recruitment. You pick up the phone and you've got a black book of contacts. And it's like, yes, true. But also we have a huge amount of tech. And it is quite funny because I've been doing this for 20 years. So it was absolutely true that it was a phone and it was a bunch of contacts. but now it very much isn't that but the problem is most of it is not holistic so people are offering you solutions and it does one thing but it doesn't talk to all these other things that we do and we don't even get me started about linkedin if you if you're if you're someone like me you will have three different systems on linkedin and none of them talk to each other um and then you then then they wonder why you're not happy when the price goes up every month but anyway let's Let's hope my LinkedIn account manager doesn't see this because it might go up more. I think as a sort of final question that I wanted to ask you is that it's great to talk about these things and efficiency and in abstract terms. But have you got any, and I don't know how much you could share with this, but real world advantages that some of your clients have demonstrated to you from these embedded offerings? And, you know, in terms of I don't know what the outcomes they were looking for, but what have they been able to kind of share with you in terms of what's been delivered by these products? So I'll talk about a few things, a few different dimensions in general. So we've we've just recently helped a startup for specialty car insurance for classic cars. It's public open road. they were able to bring up our complete suite along with a digital front end for agents. I think it was less than six months. So they're backed by private equity. They're able to bring their new products to market into multiple states, and they're going to have a continued rollout in that. So that's really around kind of speed to market aspect. And we've done that with existing carriers who want to kind of stand up a new line of business as well. So it doesn't just have to be a startup. I think then you look at some of the things that we're able to do as it relates to, say, loss control, working with some large property insurers, you know, in different parts of the world. They've been able to identify risk that maybe they were unable to see prior to using the loss control solution that we have. Because we've got a database of properties that have been analyzed, hundreds of millions of data points that we're able to kind of look at that. And it can really assess that much greater. And then you look at, say, the embedded intelligence. You've got, say, for example, the first draft of refining a claims notes, you'd have a 94 % improvement in efficiency. Or if you look at, for example, billing, you've got changing the due date, 72 % efficiency, putting a policy under notice, another 72%. Or if you look at policy, you can do a cancellation or reinstatement at 88%. All of a sudden, we've actually measured this stuff. You've actually been able to do some things at a level of efficiency. So instead of thinking about replacing legacy and getting rid of the old technology and the costs associated to that, think about how it can change the business, how you can get products to market quicker, how you can better assess and manage risk or prevent risk. Even with loss control, you can use those loss control reports to be able to go and work with the customer to say, if you did this and this and this, you could reduce your risk. So you're helping to prevent the risk. Or from an underwriting standpoint, how you can better underwrite using some of the AI models. I think that the opportunity there is a pretty broad perspective. So when customers, in fact, I was just on a call last week with a customer, potential customer, and they're trying to put their ROI together, you know, their case for why they need to replace a piece of core system. And they were doing all the traditional things. And I started walking them through, think about the productivity from embedded gen AI. Think about the AI ML models that are embedded. Think about access to your data. Think about how it could actually help improve your AM best innovation score. And then ultimately how that can impact your operational costs, your combined ratio, your growth, all of those types of things. And all of a sudden they kind of begin to kind of change their mindset as to how do I kind of justify this? So it's not just about reducing cost in one area. It's about a much bigger picture. Does that make sense? Of course it does. Of course it does. But you're a brilliant salesperson. So I don't know if I trust you. I am. everybody's a salesperson i know when i started my first business and and and i my my mom this is my i was in my 30s but my mom still insisted i spoke to my uncle because he's a very successful businessman and he went alex as soon as you know that doesn't matter what you do as an industry if you're leading it you're selling it all the time so uh yeah no and i always mean that as a compliment so look today if you've been really generous with your time i've wanted to do this for a long time. So thank you so much for spending time. Thank you. I know you're a really busy person. And you're the only person that seems to go to more conferences than me. So I'll probably see you at the next one, which I don't know. ITC, are you going? I presume you are. Yes, I'm going to ITC and I'm going to be going to the NAMIC annual as well. So a few conferences coming up. Great stuff. Well, I hope I get the pleasure of seeing you in person again. Thank you very much for being on the podcast. Thank you.