AI and the Future of Insurance with Travelers Chief Technology & Operations Officer Mojgan Lefebvre

Absolutely. So I'm Chief Technology and Operations Officer at Travelers Insurance. I've been here for about six years now. Prior to this, I was at Liberty Mutual as CIO of the commercial business. I was there for about eight years. Before that, I actually have a different kind of mix of experiences. While I started out as a software engineer and worked for a couple of years, I decided to go to business school to get away from it, but now clearly have ended back up here. Right out of business school, I joined Bain Strategy Consulting, did that for a few years, and then got back into the world of technology at a startup doing M&A and strategy stuff. But then I ended up mixing both my consulting background and my IT background and started my own small IT shop. Got recruited by one of my customers to start their offshore development, which I did in Argentina. Was promoted to their International CIO, and that kind of started me back into the CIO role, which is really, I would say, a business leadership role leveraging technology heavily obviously. So I became CIO at the company I was at, which was Teletech. From there, was recruited to join a French healthcare company, Biomerieux, which was in France, and so I was going back and forth for about four years. Got really tired of that commute and then got recruited into Liberty Mutual, which was local to Boston, which is where I'm from.

Yeah, absolutely. So we've been around for over 165 years. We're an insurance company, P&C property and casualty insurance. We're a top 10 property and casualty insurer. We're the only P&C insurer that's on the Dow Jones. We have over $40 billion—this year we broke $40 billion in revenues. Have over $120 billion of assets in our ownership. We've got over 32,000 employees, and we're number two P&C commercial insurer. We're in the top 10 personal insurance, and then we're the number one workers' comp insurance company.

Well, you know, we don't manufacture anything; we have no physical goods. Everything we do is really based on data and information, and so insurance companies have had systems and therefore data in those systems for years. And really, how you leverage technology to get insights from that data is core to how we do business. We write risk all the time, and so we've got to make decisions, and our ability to make those decisions in as informed a way as possible is really critical, and our ability to predict as much as we can, to segment the risk, to price it as accurately as we can. So that means we have to be able to store data, make sure that the quality of our data is as good as possible, and then the ability to really take value out of that. And then of course, in today's world and the consumer-based economy that we live in, really everything from customer experience all the way through making sure that we're providing good pricing and quality to our customers is key. So you can actually use your mobile phone to do everything from getting a quote for your insurance, looking at your policy, downloading it, looking at it. If you're in an auto accident, you know, taking a picture and really doing everything online and that first notice of loss that you give. So there's a lot of digital capability embedded in that. That's one area of focus from a technology perspective. Secondly, in that risk arena that I talked about, of course there's a huge amount in terms of underwriting and claim and triage of that claim and designing it and evaluating the severity of it. So there are many different areas across the entire value chain of insurance, from acquisition all the way through claim and servicing, that we use technology every step of the way.

As you said, insurance companies are probably one of the first, if not the first, big data companies. And so the ability of AI to leverage data, and if done right it can truly transform a lot of things. So we think that this can be huge for us, a company that has so much data, so much of which is curated, whether it's our claim documents that we have that we use to train our claim professionals, or our underwriting standard operating procedures. You know, all of that knowledge. Now the ability to bring all of that together and to leverage a technology that can go through that so quickly and put it at the fingertips of your employees to ask any questions and really get that knowledge directly and easily and as fast as one can imagine, as opposed to having to go through multiple documents. So I think that's just the tip of the iceberg for us, and we see huge opportunity, and that's why we're very, very excited and very focused certainly on AI and generative AI.

Going back to the first principle of saying anytime there's a technology, we shouldn't be saying 'OK, let me go look for a problem' and starting with what's the business that we're trying to solve, what's the objective we're starting with, and then does AI come into play? And then thinking of the technology tools that are at our fingertips, and AI and generative AI being one of them, a very vast field, but just one of the many that you can use. And then implementing—I think that's how you can really get to where it's going to be real. And I think that it can have huge impact. And making sure you don't fall prey to just saying 'I'm going to roll Microsoft Copilot to everyone and I'm sure that we're going to have this much productivity come out.' It just doesn't really happen that way. Being thoughtful, being selective. We've come up with a thousand use cases doing a lot of workshops across our entire enterprise, but there's probably less than 10 that we're focused on because we want to make sure we're focusing on the ones that are really going to make a business difference that we can scale. There are some where we say we're doing this for learning, and of course everything is new, so obviously you're going to make mistakes, but you want to make sure you make them the right way, you fail the right way and learn from it, and then continue to iterate. Of course, unless you start experimenting and doing and being thoughtful about the business outcomes you're focused on, you won't really get there.

Sure. I think I shared one of the more basic ones which sounds basic, but again it's got the potential to save hours and really take away the drudgery out of the work that a lot of your employees have to do. And that's where you've got knowledge and documented, curated knowledge that's in the form of many unstructured documents. The ability to leverage that is like a base, and I think that's a lot of companies are really working on that. As I look at the specific use cases that right now we're investing in and building, one of them is our interaction with our distribution partners, so agents and brokers. Especially in the world of commercial insurance, but for us also on the personal insurance side, they play a huge role, and they're really the interface between us and our customers upfront. We actually believe that there's huge opportunity there to really cut down the time that it takes to go from where all of this information comes in to getting the right ones in the hand of the right underwriter. So really that submission triage upfront is absolutely one key area. Then in the world of claim, certainly there's huge opportunity as well. Again, when an accident happens and if there's an impression that potentially there might be fraud involved, in fraud detection I would say there's certainly an area where you send your investigators to investigate, and they've got to again go through dozens of documentation and research that they have to do. So again, the ability to bring all of that together, have your computer, your AI do all of that work for you upfront, it again cuts down the amount of time that you have to do upfront to get to that fraud detection, which frankly every minute and every hour and every day that you save really makes a difference. And then last but certainly not least, I'd say personalization and the customer experience. So really, our customers expect for us to know them and provide service to them the way they want and where they want it. So again, leveraging AI, not necessarily just generative AI, but leveraging AI and AI-assisted service to make sure that we have all of the information about who they are, their preferences, everything that we can, by the way, glean from already existing materials without having to ask them the question. All of that is really important. So for example, when you want to get insurance for your home and if the agent asks you what's the shape of your roof, most people wouldn't know that. And this is information that actually either exists out there, or we've built the capability along with some geospatial capabilities to actually know what the shape of your roof is. So these are the type of questions that can be pre-filled and people don't have to even ask. So it really impacts every part of our value chain.

Insurance is such a different business in terms of the interaction that you have with the customer. A lot of that interaction is actually in moments that aren't necessarily the best moments of life. So to the extent that we can make those moments less hectic, with the least amount of anxiety and as seamless as possible, that's really where we want to take it to. And so we're very focused on seeing where we can really have everything be done in an automated way and in the easiest possible way. So as we think about it, we think about the fact that if we think of the claim experience, how can that be as straight-through as possible so that it's as digital as it can be? But we also have to make sure that we provide the ability for the customer to call a human and have the human be there, because you want to make sure again that it's an omnichannel experience, that that choice is made by the customer and it's not us. And then in turn, what that will do of course over time is to really make you efficient as an organization, how you run internally. And the more automation and intelligence that you're using to get a lot of these things done, then the more you can have your people where it matters. It's those human elements that AI, hopefully, never have that are the areas of focus from a skills perspective that you need to focus on—your ability to have empathy and really be kind to people and these kinds of things.

One thing is, while I think there will be huge impact over time, I don't think it's right around the corner because I think until—and you said it—until people start reimagining processes and the way they work and then leveraging technology towards it, they won't really get the full value out of it. And I think that's not something that you can do overnight. And then, as you said, you certainly will need more people who know how to maybe not necessarily create AI, because again there are companies that are very focused on that that understand how to leverage it already. But we do know that of course it's very important how you ask questions and how you really prompt your AI. I think there's actually a whole series of roles that get created there. And I actually think it's going to continue to be as important, or more and more important as it is now, but it's going to become even more important for your IT organization to truly understand your business and your business processes and have the ability to think of where these technologies will actually make a difference. And you can't really do that without understanding the business process and where you can inject the technology.

We're lucky that we have a CEO who is very forward-thinking. And he uses the saying that at the point where the pace of change outside of your company is faster than the inside, then you know the future won't exist. And looking at that and the impact on the changing technology, the importance of data and analytics, all of the forces of change that are important in our industry, the impact that that has on our distribution, which again, as I said, our distribution partners are a big part of our value chain and very important. So he understands that. And so with that, while we've always focused on our performance and the return on equity and the return to shareholders, which I will openly say is critical and important to us, we've also established our innovation priorities. And it's very simple, and I bet you that every single one of our 32,000 employees can say what they are. It's about great experiences, advancing our advantage in risk expertise, and thirdly, creating productivity and efficiency. So these are our innovation priorities. And so everything we do, we've got to have the ability to attribute it to these three. And this in itself first of all makes it easy for each employee to know where we're going as a company and why what we do is important, and then how their work relates back to that, and gives them the right to say, 'Well, what can I do in my job to really change it?' And then some of the things that we've done in the technology organization is really try to see how we can really make this real. And we've done things that are both bottom-up and top-down. So while we consciously, as the leadership of the technology organization, decided that we would continue to make sure everybody understands what our business objectives are and thinks about that first and foremost, and I open every single town hall that we do on a quarterly basis with starting with those business objectives and then drive down to well how does technology really impact that. The other thing when I joined was there was also a tendency for each area to want to create their own solution. And so I knew that really coming in and ensuring that, for where it makes sense, for us to build platforms, reusable platforms and components that can really be leveraged across the board, and that everybody doesn't have to build their own thing. So the motto that we've created, which again is not only used in the technology organization but across the board, is 'Common as possible, unique as necessary.' And that's really become the way we operate. So where are the horizontals that you can do common, and then of course you've got your verticals where you can be as specific as you need to be in a world where everybody expects personalization. And then on an annual basis, we have an innovation jam, what we call an innovation jam, across the entire business. It's not a hackathon just for tech, but the entire business does it, and we allow anyone to participate. Teams come up with ideas that are always aligned to any one of the business objectives that we have. And this year, what we did was we made gen AI available, and it was incredible. The majority of the ideas that came out were absolutely leveraging gen AI. So those are just a few things, Evan. But then maybe let me also say, you've got to model the way. The idea of being open, being there, being inclusive—if those are just words that people don't believe, it's not effective. But just earlier this morning, I did one of my monthly coffee chats that I do with our new people who joined the organization. So making sure that you're continuously talking to people regardless of where they sit in the organization, and that they all feel that they're part of the overall Travelers mission, and the fact that we help people live safer lives and really deliver on the promise that we make to them becomes part of their DNA.

Business outcome.

I think every job is all about relationships. If you get everything else right but don't have the right relationships and the right connections, you won't really succeed in your work.

As a business leader. Make sure you understand their business. Make sure that everything you do about your technology investments starts with their business objectives in mind, and the rest will come.

Radical Candor by Kim Scott. It's about how you behave towards your employees and how you are as a leader, what kind of a leader, manager you are. And she's got four quadrants, and she really highlights the fact that the only way to be a good leader is making sure that you're developing people and doing it in a way where you're giving them very candid feedback, both positive and, if needed, constructive. But doing the constructive in a kind way and, you know, it's just alone but doing it in the moment. And that's one of the hardest things for any manager to do. So it truly highlights the criticality of the role of a leader in developing individuals, and she just does it in such an amazing way. So love that book.

I have yet to buy the Oura, which is the ring that tells you how you sleep and all that. I meant to do it, but then we were going on a trip to Africa where we were going to be away for a few weeks, and I was like, 'Oh, it's going to arrive just as we're not there.' So I'm very excited about it because our CEO actually instilled in the entire company how important it is to sleep and the book Why We Sleep. And actually, in our senior leadership meeting this year, he had the author of that book come and talk to us. He's a scientist, the writer of this book, and very scientific-based. So I'm excited about that technology and hope that it'll help me sleep better and give me better health, so to speak.