Travelers CTO/COO Mojgan Lefebvre on the Intersection of Technology & Operations | Technovation 782

Hi Peter, thank you for having me. Always a pleasure to speak with you.

Absolutely, we are a property and casualty insurance company, one of the top insurers headquartered in the U.S., operating globally in the United States, Canada, and Europe. We have around 32,000 employees, work with 13,500 agents and brokers across the U.S., and $37 billion in annual revenues.

Certainly. So I have global responsibility for everything technology, from anything that touches our customers, agents, our brokers, and all the way back to all of the back-end engine that runs our business. And then from an operations perspective, it's really customer service from the moment that someone decides to ask for a quote to actually getting that policy and the binding of the policy, and then any service that's provided thereafter, except for claims, which is part of a separate organization. From a technology perspective, Peter, I'd say it's everything from data and analytics to cyber to of course infrastructure, our architecture, and all things digital.

Yeah, absolutely. And I'd say it's probably more prevalent in financial services, as you look at that, and perhaps it's because everything that we do and touch is about information, and all of our operations really run on data and insights. So the more I've been in the role, the more I've appreciated why it truly does make sense. And as technology leaders, we always talk about the fact that before being a technology leader, really a business leader, understanding the business, understanding who our customer is, what their problems are that we want to solve with technology, is absolutely critical. And I'd say having responsibility for the operations, which is really where we touch the customer more than anything, has been phenomenal because there's just a natural ability to be connected to that part and really understand the customer really well, really think about experiences and what they go through, and come at technology and our strategy that way, really always focusing on what experience we are creating, how we are going to delight our customers even more.

So at the highest level, it is discrete from the perspective of how we think of the functions. So I have one leader that actually leads operations globally across, and of course has teams that sit very closely in the business, and then we have enterprise teams that cut across and really serve horizontally. And then from a technology perspective, again, I have four CIOs - market CIOs, business units CIOs - and then the rest of the leaders again are horizontal capabilities that go across. So the way I think about that is this is the functional view, Peter, but then what we've done explicitly, and this has been a journey that we're continuing of course, is really getting our teams, especially where we're building products and solutions, to sit together. So meaning, while you may be part of a what we call a chapter or function, you're part of an agile team that's doing work. And so we have structured ourselves as teams and then teams of teams. At the highest level we call it a value stream, and our value streams are focused on our customers, our agents, and our products. And so that's how the groups come together. And having functional responsibility for the two, it's just been much easier for me to really drive making sure that it's okay for teams to sit together and having dedicated operations professionals who act as the product owner within each of these teams, because they really are the ones who understand what they're looking for.

Absolutely. So I would say first and foremost, of course, we want to be as aligned to the business strategy of the company and the critical priorities. So that's first and foremost. And so we have an annual strategy meeting and a couple of days that we spend together as a leadership team along with our CEO, where we talk about the strategies that we have and how they're being adjusted. Because we establish the strategy once every few years but continue to adjust it as we go through the years, and we adjust our multi-year planning as well. Then from there, we really take a look within the technology and operations organization, but really probably even more specifically within technology, to see what are the problems that we're trying to solve and how we are going to address that. The one thing that was very clear from the beginning to me was that for us to be able to leverage our scale, we've got to do things as common as possible and as unique as necessary. And that's almost become a motto for not only the technology and operations organization but the company as a whole. And so in order to do that, I've established these what we call platform teams that are really working on capabilities that can be leveraged across the board, and then capabilities can be built on top of them. So platform teams like our sales and service enablement is a platform, our product and policy is a platform, again all of these have a foundational element to it. Our cloud platform is another one. And just think of those as platforms that are being built like products or platforms, and these teams have to think ahead of what the business is going to need and make sure that they're in place before the need is there, or at least in good time when the need is there. So for the strategy to come together, there's a lot of conversation and planning sessions that we have among the horizontals and the verticals. We probably get together three times a year at least with this leadership team and some of their extended team members to make sure we're figuring out what the dependencies are, if we have shortage of resources in any of these areas. So there's a lot of collaboration on planning that goes into that.

Absolutely. So Peter, I think of it in four very large buckets. First and foremost, it's really focusing on creating great experiences and capabilities and platforms for our customers, our distribution partners, and our employees. And I'll come back and give examples to that, but that is really the crux of what we want to do, and we want to make sure that we're impacting business outcomes. So that's core to what we want to do. But in order to do that, there are a couple of other things that we're very focused on as well. I'd say that second bucket is really modernizing, and the modernization is focused on two things: one is our technology architecture and infrastructure, so really moving from the monolithic kind of older type systems to everything being API-driven, microservices-based, based on eventing for data and analytics. And then I would say the second one is really driving engineering excellence and continuing to make our engineering practices better and better. Then the third one is really putting insights and data into everything, and you and I have talked about a lot of the efforts that we've had for many years now leveraging machine learning, leveraging the data that we have, our own rich data and the third-party data. And then of course fourth, I say it's really protecting the organization, and really what we think of as absolutely cyber is critical to us, and this is also where our business resiliency responsibilities lie. And it's beyond just technology or just operations; it's really business resiliency for every business process across the business. So these are the four areas.

Absolutely. So some of that of course comes from our business partners and the interactions that happen. And on an annual basis, by the way, we have what we call the leadership conference where we get our top agents together. I always participate in those, and we really have sessions where we try to understand what they're looking for, what are the problems we've got to solve for them, etc. We've also got a research team that does a lot of studies to really make sure that they come back with the right themes, whether it's through focus groups or other ways. They really have expertise in coming back with what the core areas are that are going to make a difference in our net promoter scores and other things. And then I say the third way is again that direct interaction that the operations professionals have, because they're really the ones who are interacting a lot both with the customers, once they've become a customer, and also with our agents and brokers that provide over 99% of our business. So we are very much an organization that works very closely with our partners, and it's a very tight relationship, and I think it's a mutually appreciated relationship.

Yeah, that's a great point, Peter. I would say it certainly probably has taken me two or three years to really get different parts of our business to the point where they absolutely understand that modern architecture is critical to what we want to do. Because what I always say is you can only innovate that much at the edge, and if you don't really take care of the core which is underneath, it's going to get in the way of your agility and your ability to really respond to the market. And I think what we've had going for us is the pace of change that's just happening out there. We did start with a big kind of education for our partners just to say what do we mean by modern architecture, and then using companies like Uber as an example to say if we want to one day be the insurance platform and really think out a few years, and we want to be that platform company for the insurance industry and beyond, we've got to be built in a way where you can do that. So everything that we do, every service, every process has to be built as an API. It's got to be built in a way where we leverage eventing for knowing when something's happening so that in an automated way, the claim can start occurring based on what data is coming to the different solutions. So really making sure that we talk about that. And interestingly enough, while in the beginning people were not as comfortable with the words, it's funny, we're now even our business partners are talking about APIs and really getting it. So that's been very helpful. And what's interesting is even the agents and brokers have started talking about that. And I think it was the pandemic that probably also drove some of this need to really understand what it takes to truly have a much more agile infrastructure, one that's far more modular than what we have today.

Absolutely. So we've again, as you said, we think of ourselves as a data company. When you're an insurance company, you're making all your decisions based on risk, and it's all based on information. We don't manufacture anything, so data is critical and core to us. And every single person understanding that the data that they put into our systems, they have to take care of that data, make sure that it's quality data and as accurate as possible. So we've really honed in on first of all making sure that every one of our associates understands what data is. And there's a training that we put together now that every single person who is part of the company or joins the company actually, it's a mandatory training that they have to take. So that's how we've made sure that people understand data. And then really again ensuring that we prioritize all of the places where those critical insights and insights into next best action are really embedded. And the one thing we've said is anytime that we go to build a system or a solution or implement a package solution, we're going to think of the data first and the information architecture and ensure that that data is going to be visible at some point across the enterprise and understandable. So really standardizing on taxonomy and terminology and information architecture, all of that for the ability to then leverage algorithms and models that we use for risk assessment, for pricing, for reserving, and now actually also for customer experience. So the ability to leverage this data through those models, whether our own or in some cases they come with a lot of the platforms that we use like Salesforce, ServiceNow, and others, the ability to really leverage that data is critical. I'll give you one example, Peter, which is more recent than the last times that we talked. Fraud, as you can imagine, is one of the areas that in the insurance industry like many others is something that we've got to be very careful of. There are actually rings where you've got bad actors from lawyers to doctors to other people who band together and submit a lot of fraudulent claims. So this is something that makes the insurance industry lose hundreds of millions of dollars, and at the end of the day it impacts all of our customers negatively because insurance companies have to raise their price to cover all of these losses. So the ability to detect these as quickly and effectively as possible is critical. So our platform team actually just created an AI model that is being leveraged with some of the more modern database technologies as well to have the ability to really visualize these patterns. So our investigation team, which is a team that's established to make sure that they're looking at things that look a little bit fishy, leverages visualization to actually see these patterns much earlier than we would have seen it if we didn't have the automation and the technology behind it. And it's been incredible - the ability to probably save three to five years in detecting these fraudulent cases. So this is just one example of how we're leveraging machine learning and the data that we have, and also leveraging external third-party data.

Absolutely. And really where we're focused right now is trying to see how and where we can leverage machine learning as much as possible to really drive as much self-service and straight-through processing, whether it's in a claim or any other area. And we've started with more of the simple cases where we've actually already gone live with those, whether it's in the property area with food spoilage or simple theft, and then really with the ambition to try to increase that. And the incredible part is that the customers are really adopting it. Customers don't want to necessarily interact with you if they can do something themselves. However, we of course have our wonderful claims professionals who are there to assist if they need to.

Yeah, absolutely. It's funny because some of our operations professionals who might have used dummy data in some fields... this wonderful woman was like using the birthday of her daughter for everything. But then we really have the teachers walking through and saying, if you're doing something like that, that's how it's going to come back and it's not really going to help us or help our underwriters in the future. Yeah, it makes sense.

I mean, I think diversity and diversity of thought and the ability to really make sure that anything you're doing, you are thinking of it from all angles, is absolutely critical. And I would say in a world where again the makeup of who we are has changed and is going to continue to change dramatically, and who our partners and customers are tomorrow is going to be very different, so if the composition of our teams is not similar to that, then we're not really going to be effective in working. So it's also a business imperative in addition to the fact that it's the right thing to do. So I am very passionate about this, I would say especially when it comes to women and their role specifically in technology, because again there are far fewer women in STEM-related roles. So really trying to make sure that we have an environment where there are more and more women who can be part of the tech organization and rise into leadership positions is important. So roughly five years ago, when I started, we established what we call Empower Plus, which is our diversity, equity, and professional network for technology and operations, to really inspire women to have long-lived careers in these fields. And then making sure, most importantly, for allies, which is what the plus sign indicates, to be part of this, because you need the people who are in positions of help to help those who aren't. So our men and non-women are as important as the starting for us. And we have several thousand members, and both our men and women are extremely active in this area. And there are many, many programs that come out of it, from mentorship, one-on-one or group mentorship, to learnings that we do, to hackathons that we conduct, and many other things.

Yeah, I mean, I'm sure Peter, you're getting the same answer potentially when you ask this question. It's hard for anyone at the moment in history not to talk about generative AI, and really what ChatGPT did to probably make the understanding of what AI is and can do far more easy to understand for anyone. It's like hundreds of millions of people joining ChatGPT and understanding how it can make a difference, and of course with its pitfalls as well. So I'd say there's huge potential there. I think it's the first time in a long time, maybe the internet was the last time, where something was seen as having the potential to truly disrupt everything around us, both positively and hopefully less negatively, although any technology can be used for the negative. And when I think of it, it's really about augmenting what we do as humans, and I really hope that's where it's going to go. It's really leveraging anything from the knowledge bases that we have across our companies, like thousands of documents, but the ability to now leverage those much more easily to build solutions that can be really at the fingertips of our claimants or underwriters or any other professionals or HR people. The ability to leverage it for code enhancement is incredible. Whether it's Microsoft's Copilot or Amazon's CodeWhisperer or any of the other ones, or OpenAI's Codex, apparently our engineers tell us this is really high quality. Of course, there are all the regulations and copyright issues to think about, but it's hard not to be very excited about this. And the beauty of it is we've got our AI accelerator which we established around three and a half years ago. We've got a phenomenal team that's really very focused, and over a year ago they were very close to what was going on with OpenAI. So we've even shared some of it with our board of directors, who was very appreciative. So I would say everyone in our company is very aware of what's going on in that realm.

Absolutely. I would say what we've made sure is anytime that we're working on a use case, we work very closely with members of our business team. One of the use cases that they've worked on and made huge progress on is again partnering with our operations teams and some of our business teams in one of our business units where in today's world still a lot of the business with our agents and brokers happens through email, believe it or not. And these are emails that come in with dozens of documents attached to them. So we've got people sitting offshore that are processing and looking through those documents and filling out forms, etc. So working together, we've built the capabilities for a lot of these emails and documents to be ingested, to be processed and understood, and really do a lot of that pre-filling in an automated way. And I think with generative AI and as it's getting better, the possibilities in this area are tremendous. Something that might take weeks to do today is probably going to come down to hours if not minutes at some point. We think we are certainly ahead of the industry, but there's no doubt that this is going to catch up and the entire world is going to leverage these capabilities. Some of the other things we make sure we do: one is focusing on use cases with business value, and what's powerful is when you bring this to the management company and talk about it, you've got your business leader or the head of field operations talking about it, it's just got much more value than just your technology teams. And then the other thing that this group does is really again making sure that they're teaching what they know and how they know to do it to others. So providing anything from study groups where they get together and literally teach people how to do Python programming to other things. Their goal is to democratize AI and to make sure that every single one of our engineers for sure is also an AI engineer, but beyond that, the tools that everybody can have at their hands, eventually the entire power of the 30,000 can be leveraged in this around.

The other thing we're doing, Peter, sorry to interrupt you, but one other thing: again taking a page out of what we did with the data culture training. I was just talking about this with our learning leader who's done a phenomenal job of really thinking about what we need to be doing, and putting together a curriculum which we call Digital Fluency for the Enterprise. It's got different modules, but there's a tech foundation module that we think is absolutely critical for every single person in the organization to take. So again leveraging both internal and external resources to pull this together. We just did a pilot with around 150 of our operations folks, and yet again, another advantage of having that group in my organization, we tried it out and the feedback was phenomenal for these 12 courses that take up to nine hours to take. They're micro-bursts of video where concepts from the API economy to cloud and other things are explained in a way where everyone can understand it.

Yeah, well thank you for that question. Of course, I'm privileged, let's put it that way. I'm lucky to have had these opportunities. When I think of how this happened, I kind of think of my journey in buckets. The first one is that I was born in a country where unfortunately due to a revolution and a change of government, it was obvious to me that women wouldn't have the opportunity that I wanted, and being very ambitious, I knew that I needed to leave that country. My parents couldn't support to send me, so after graduating from high school, I actually worked for about a year and a half, put some money together, and applied to schools in the US. That was my desire to come to the US, and I was lucky enough to get the visa. At the age of 19, with just a few dollars in my pocket, on my own, thousands of miles away. So when I think of that, I think that took courage. So courage was absolutely a big piece of what has driven me to where I am, really not being afraid to stretch yourself and put yourself in situations which may not necessarily work out the way you thought, but if you want to do it, just go for it. Having that courage is absolutely critical. And as I think through my career, those decisions have happened multiple times. I'd say secondly, Peter, I tell everyone including my kids that one of the most important skills to have regardless of your role is communication skills. The ability to read and write well, and to communicate in a way where regardless of who you're talking to, you can have the conversation in a way where you're telling the story in their context - that's absolutely critical. So that has certainly paid off. And then my third C: courage, communication, and my third C is connections. I would absolutely never have been able to be where I am without a sense of connections and relationships. Some critical connections that at the end of the day, without the help of others, you can't really progress. And that's where allies are so critical. When I think back to my career, probably one of the biggest drivers of my success and the ability to go into leadership and management was one manager that I recall where he and I bonded around talking about cultural topics. He was from India, I was from Iran. He took a chance on me and gave me the opportunity to become part of a project that became a very global project. I think that really propelled me into leadership where it might have taken me much longer. So those three - courage, communication, and connections - have really made a difference in my life.

Thank you, Peter. As always, a pleasure. Thank you for the conversation. I always learn so much from you. I appreciate that. Thank you.