In today’s rapidly evolving technological landscape, Artificial Intelligence (AI) and generative AI (GenAI) have emerged as pivotal tools driving transformation across various industries. Nowhere is this more evident than in the insurance sector, where the integration of AI promises to revolutionize processes, enhance customer experiences, and optimize operational efficiencies. However, amidst the excitement and buzz surrounding AI, there is a crucial need for a deeper, more strategic approach to its implementation.
This was the focus of a recent conversation on the Insurance Unplugged podcast between host Lisa Wardlaw with industry thought leader and strategist Denise Garth, Chief Strategy Officer at Majesco, the cloud insurance platform software for insurance business transformation.
In this episode, the practicalities of AI and getting AI ready take center stage: the importance of a robust data strategy, the impact of legacy systems on your data, and why tackling the challenges of the future requires a new approach.
Here is an excerpt of their conversation.
Current Perception of AI
Lisa Wardlaw:
How do you see people thinking about AI and are they thinking about it broadly and deeply?
Denise Garth:
As we both know, GenAI and AI are the hot topics of the industry at the moment. And sometimes when it’s a hot topic of the moment, everybody wants to say, oh, I’ve got it, I’ve got it. And do you really have it?
I think there is a danger, quite frankly, that people are reacting to what’s hot out there. AI has been around for a long time. We’ve got AI models that have been used for over a decade or more and new ones that are coming out. Those are really great, and I see that those can be point solutions because you’re kind of focusing in on a specific problem like fraud or maybe a specific type of risk for underwriting or maybe something in claims like subrogation. But you also have the danger if you don’t have an overall data and analytics strategy that says, ‘how are you going to put all of this stuff together?’
And most importantly, where the heck is your data? Do you have access to all of your data? Is it clean data? Is it valid data? If those things aren’t in place, those point solutions or even a more holistic solution is going to be very, very difficult.
So, you’ve got to have an overall strategy. You’ve got to be able to pull all your data together. You’ve got to make sure that it’s valid and clean data, and then what are going to be the tools or the technologies you’re going to apply to it?
From a GenAI standpoint, that’s even more important than ever because I personally do not see Gen AI as a point solution. I think it minimizes the potential of GenAI tremendously. You’ve got to look at it across the entire process, end-to-end, from an insurance perspective. If somebody says they’re doing GenAI in a point solution, I have to question, what’s going to be the value of that and how are you getting the data? Where are you getting the data from?
The Role of Data in AI Implementation
Lisa:
My hope for taking GenAI outside the hype and into real practical application is that we leverage it as a catalyst to say we can’t do digital in a point solution manner anymore. Basically, it’s going to require more holistic thinking to make it value added. How would you connect that idea of catalyzing the data, for example in the Life or Annuity space? Do you see people thinking that way yet?
Denise:
It’s in pockets. And I think that unfortunately for the life side of the industry, it’s even more of a challenge because we’ve got such tremendous legacy. I mean, when I was in one insurance company, I had 11 different admin systems. Those legacy systems were highly customized. I say that they were bastardized because basically we define fields differently to be able to do different processes at the time because the technology wasn’t as robust as it is today, to be able to quickly add additional fields to do configuration, to do the things that you can do in these modern solutions.
And so, a lot of the data—sometimes it’s embedded in the code, or the definitions of the data are embedded in the code, and then you don’t even have consistency in the definitions. Because we’ve got a lot of business running on those legacies, they may have put up a new policy system to start some new lines of business or to bring out some new products. You still have all of the old stuff. And sometimes that old stuff is valuable because it has the data insights, because it’s got the years of history, particularly for life health, the years of history is really, really important.
So, the ability to pull that data out becomes really important. That’s another aspect. I think even a lot of modern systems don’t have the ability to allow insurers to pull all of their data out of it. You only get pieces of transactional data. You don’t have access to all of your data, and if you try to get access to it, it becomes a really huge kind of piping engineering thing that becomes really, really expensive.
I think that one of the really critical elements here is that what has really evolved, I think in the last year or two, is that the technology foundation of your systems, the architecture, is absolutely critical in making decisions because it’s going to allow you to do things or not allow you to do things. Things like a native cloud, a three-tier architecture, like being able to have access to all of your data APIs and to your point embedded analytics. Those things become critically important because they become the levers that allow you to do things uniquely for your business, but you have access to the data that you don’t have. And I think finally people are beginning to kind of click on this and think, oh my gosh, I don’t have access to all of that data. What am I going to do?
The Need for a Robust AI Strategy
Lisa:
If you are going to build an AI strategy, you are going to need this data very quickly and the data that you’re typically receiving is the transactional output data. The data that you need to power AI is what I will call the foundational data, core data. How would a company know if they own their data? What should they ask? What should our listeners ask their technology team or their vendors?
Denise:
From a transactional standpoint, you’ve got data coming in that gets massaged and processed, then you create data, and then you’ve got data coming out underneath all of that. Do you have the data elements, the data definitions and the data structure? Do you have access to all of that? Is it defined? Can you easily add data elements into your data structure? Does it have an easy way that it will actually pull it out into a lakehouse, just pull the data into a lakehouse and allow you to add additional data easily? Does it allow you then to massage that data, to pull pieces of that data out and use that as part of training an AI or machine learning model? Are you going to use those pieces of data? Maybe I’m pulling some from policy, maybe I’m pulling some from claims and maybe I’m pulling some from CRM and I’m going to do a BI report. Can I do that on top of it?
Underneath it all, you’ve got to have access to all of the raw data that’s coming in that’s being created in the processing and that any other data that you want from a third party standpoint, and that’s typically defined within a data model or a data structure of some type that you know what you’re pulling out. It’s the raw definitions of data.
If you don’t have that and you don’t have access to it, you’re going to be limited in what you’re going to do. So, all you’re going to get is what we’re going to show you in reports. That’s the data you’re going to get. But that’s a very different type of view of data because that’s basically operational reporting to say, what am I doing? It’s not allowing you to do the ‘what ifs.’ Well, what if this, what’s this going to tell me if I did this, what’s this going to tell me? Those are the most important things from an intelligence standpoint that we really haven’t had access to.
Insurance Industry Call to Action
Lisa:
What would you like people to start doing? What would you like people to stop doing and what would you like people to continue to do? What would be your call to action for everyone listening?
Denise:
I would like people to start by seriously putting together a data and analytics strategy and doing that means that you’re going to have to really, really look at your underlying solutions and technology and architecture. Is it going to get you where you need to go and be able to do that?
What I would say to continue doing is that once you’ve kind of done that good assessment, what initiatives do you have underway that will fit into that? Keep those going. You may have to tweak them to make sure that you’ve got the ability to be able to pull the data together, all of that.
And then what I would say to stop investing in your old stuff, reallocate your resources to the things that are going to make a difference in that assessment.
You’re going to know, for example, that my policy system is never going to allow me to get the data I need. I need to really rethink this whole thing. I need to reevaluate it. Then reallocate your resources to that and quit investing in keeping that old policy system going and adding a new product to it. At some point, you’re going to have to cut the cord and you’re going to have to move it over. You’ve got to make tough choices and set those priorities.
I always like to say that you’ve got to really have a strategy in place. Your overall business strategy, the data and analytics strategy, has to be a part of that overall business strategy. Then, you need a technology architecture strategy that is going to play it all together. And then you can select your solutions that are going to meet the business strategy element, to support the data and analytics strategy and are built on the technology strategy that you want the business processing.
You’ve got to be able to pull that together, reallocate resources, reprioritize the most important things that will get you there as quickly as possible. You can’t do everything at once. That’s just not feasible. You’ve really got to have a plan. And quite frankly, the most important part of strategy is execution. We talk a lot about strategy, but so often we don’t execute, or we execute to one point and then we stop, and we move on to the next topic. And that’s quite frankly one of the reasons why we still have so much legacy sitting in the industry because we started a transformation project. We got phase one done, and then we went off to the next thing and we never pulled over all the old. And that’s where legacy comes home to roost.
Lisa:
What is your final call to action that you want everyone in the industry to take away from this?
Denise:
My final call to action is that we are at a crossroads in this industry. I think the last couple of years have really intensified and exposed the challenges that we’ve got as an industry with our underlying business operating models and our underlying technology foundation. It’s not going to get any better. And so, we’ve got to get very, very serious about what are we going to create for a future.
For us as an industry and as individual businesses, we need a different operating model that’s going to process business and that’s going to differentiate and engage our customers, but we need a different technology foundation to do that. You’ve got to do both of those hand in hand, and you’ve got to have a plan in place. You’ve got to have a strategy in place, and you’ve got to start. Now.
There are a lot of companies that have already started and they’re well down that path, and they’re going to actually be able to accelerate what they’re doing. They’re the ones that are going to be able to use generative AI. They’re the ones that are going to be able to use new sources of data. They’re the ones that are going to be able to use the new technologies that are going to be emerging. We know there’s going to be more coming.
Those that aren’t are going to be left in the dust, are going to struggle very much keeping any kind of market share and be in any kind of profitability. And they’re going to be shrinking businesses.
And so, we’re at a crossroads of making decisions about our future. And the future is very different than what the past is. And so, that’s why the operating model has to be different. Our buyers are different. Our risk environment is different. Our technology environment is different. Everything is different. And we have to rethink how are we going to do business in the future. So that fundamentally we go back to the purpose of this industry is to make sure that people are covered in time of need and we make them whole and that we can provide that risk. We can’t just be increasing costs that people then decrease the amount of insurance that they have, or they can’t even afford the insurance. That’s not a way out of this problem. We have to find a different way out.
Conclusion
In conclusion, the call to action is clear: insurance professionals must prioritize strategic planning, invest in modern technology infrastructures and commit to continuous improvement in their data and analytics capabilities.
Listen to this episode of Insurance Unplugged with Denise Garth, Chief Strategy Officer, Majesco.
