Olivier Lafontaine:
I'm Olivier Lafontaine, and this is Life Accelerated, the podcast for life insurance leaders focus on driving meaningful change through technology, process, and partnership. In this episode, I'm joined by Abhi Bakray, lead client partner, strategy and transformation at IBM.
Abhi has spent more than 20 years helping life insurance companies navigate wave after wave of transformation from core modernization to digital to AI, and he's convinced the pattern never really changes. It was never really only about the technology itself. We'll explore why he believes the next competitive edge for carers is shifting from products and modernized systems to something he calls open enterprise intelligence and why he thinks the old rule of fix your data before you do AI has it backward. We'll also get into the future of the advisor, the untapped growth sitting in every carrier's Inforce book and what will separate the leaders who thrive in the next five years from the ones who fall behind.
Let's get into it.
So welcome to the show, Abhi. It's nice to have you here. As we usually do on the show, we like to talk a little bit about hobbies and stuff that our guests do outside of work, outside of life insurance, outside of AI. So let's start with that. What do you do on the weekends, Abhi?
Abhishek Bakre:
Apart from playing in basketballs with my kid, I enjoy cooking food in general. So weekend is sort of in therapy time and cooking all sorts of things in the kitchen.
Olivier Lafontaine:
Yeah. What kind of dishes do you cook best, do you
Abhishek Bakre:
Feel? Growing up as an Indian in India, obviously a lot of Indian food, but also Chinese and Italian, whatever we all have grown or come to light today. It can be a burger sometimes, so it's really a melting pot.
Olivier Lafontaine:
Awesome. All right, so if we get into it, you've spent more than 20 years helping life insurance companies through waves of transformation, core life systems, modernization, redesigning, operating models, enterprise AI. So if you look back at that experience, can you share some of the lessons learned from those transformations and what part of that might still be relevant today?
Abhishek Bakre:
If I look at my career in the last 20 years, and I've at this point seen and been through all sorts of waves from core modernization to digital to AI, now in between there was blockchain, we had portals, data. And in my mind, if I reflect back as I was preparing for this conversation, it was never about technology from a transformation standpoint. It has always been around can I really redesign the business operating model around it, whichever new technology or data platform we are introducing today it is AI. And every wave across those key technological revolutions in my experience has basically exposed the same organizational challenges, but in different ways. When it was policy admin systems, it was all about how do I drive more out of the blocks workflows and get people adopted and trained on it? And now if I look at AI, it's all about how do I change human mindset and decision making around AI and what to trust versus not what to trust and when to probe versus when not to probe.
In a sense that AI isn't necessarily replacing modernization plainly, it is continuing to expose two decades of what modernization has failed to solve inheritantly with a lot of life and annuity carriers. And as we move into the next era and go deeper within AI, I think the ones that really redesign the decisions around AI and the supporting operating models, not just the systems or the data platforms will really stand to benefit the most in my mind.
Olivier Lafontaine:
Yeah, that makes a lot of sense. I should say technology by itself isn't really what makes the biggest transformation. It's the business redesign around it that makes the biggest difference in the end is this. So if you tell us a bit more about IBM and where you sit in the company, how the company is structured and what's your role, can you talk a little bit about that?
Abhishek Bakre:
So IBM, I mean everyone knows IBM as long as people know technology, you know IBM because a lot of the modern day technologies in some shape or form connect back to IBM one way or the other in the course of last 120, 30 years we've been in the market. I sit within the IBM consulting, which is in my mind a little bit of best kept secret, if you will, because IBM is three distinct organizations. You have the IBM technology, which is dealing with all things hardware, software. You have IBM research, which is really focused on new edge, cutting edge, next gen research around quantum AI and things like that. And then you've got IBM Consulting, which is our services business, a big part of our overall P&L. So I sit within the consulting business and even more so within the financial services and insurance leadership team, if you will.
A couple of different roles that I play from a firm standpoint, I lead our relationship with one of our largest life and annuity carrier clients, but also responsible for driving a lot of our go-to-market around life and annuity as well as group insurance. And that takes several forms in terms of market evidence, market building, market making, and external presence and thought leadership, if you will. In terms of my pedigree, in terms of the client challenges that I solve for or strive to solve for through various projects and engagements really is working with life and annuity executive teams, really connecting business strategy, operating model, technology, data, AI into one singular transformation agenda. Many times you have folks that do one or the other well, and the real value comes from treating it as a end-to-end transformation, if you will. So really tying from strategy all the way through execution rather than a technology transformation or an operations transformation or a growth strategy or a cost takeout type engagement.
And that is really rooted in bringing not just our consulting capabilities, but also the deep engineering vigor that IBM is known for combined with research and our partnership and strategic alliances to really end of the day drive value for our clients.
Olivier Lafontaine:
That makes a lot of sense. I guess you meet a lot of executives from life insurance companies talking about what their challenges are and where they're going, I suppose, both from a business design or organization design perspective as well as from a technology perspective obviously. What are some of the things that you hear the most or that are top of mind for these executives in the space?
Abhishek Bakre:
Sure. So whether that's a mutual carrier or a publicly traded life and annuity carrier, the conversation in the last several years, or I would say rather last five years and even more so recently has really shifted from not just cost reduction, but how would you tackle growth at the same time? So how do you drive profitable growth for the organization is key. On the distribution side, you obviously always said distribution economics play a huge role in life and annuity, especially when you're talking about a fee force and advisors. And that continues to be a key priority for most of the clients that I work with. And then taking the customer engagement to the next level in terms of how do we now solve them better, proactively service them better is top of the house agenda. And AI now is mainstream as we know, and depending on how you define mainstream, but it's really a transformation agenda both at the C-suite but also at the board level and no one's treating that as a mere technology revolution, if you will.
And on the same lines, one thing that almost every single executive that I talk to, it's clear the finding of the micro AI use cases in underwriting or claims or servicing, we always knew that. Every executive knows that where you can drive that efficiency. It's really deciding what business the insurer wants to become in the future. And the ones that are making some real progress, we are far from it yet, the ones that are making progress are really anchoring on making less technology decisions, if you will, but really around my earlier point around broader business design decisions on how they would operate in the realm of AI.
Olivier Lafontaine:
Right. And we hear that a lot. There's a lot of conversation. It's one thing to have an assistant AI and continue to operate the way you've always operated and then the truly mature or the companies that will probably succeed, I don't know if you agree with this, but at least I feel that the companies that will truly succeed in the age of AI are the ones that try to understand the technology from the perspective of how does it transform the way we do business entirely, almost rethinking how to conduct business and then deciding what would people do and what would AI do? And do you agree with that? Have you seen a little bit of that in action?
Abhishek Bakre:
Yes, I'm seeing a lot of that in action and I'm actually leading some very interesting work in that space where with one of the largest carriers where they're taking an truly end-to-end quote through claims re-imagination style approach in terms of how they are reinventing and really where they want the business to go in the future. So yes, wildly, yes. And I think that's the way to do it in my mind at least.
Olivier Lafontaine:
Out of curiosity, does it trigger some fears, I suppose? How is this topic treated in companies? Obviously when you start thinking like that, then some people's jobs are either going to transform drastically or be eliminated, potentially be transformed into something else. But I've seen this a couple times actually, but a lot of people will feel a little bit threatened. Do you notice that? Is that something that happens you feel in general?
Abhishek Bakre:
Yes, that happens quite a lot. And even in the example I used, even when the CEO and the president is trying to drive this unprecedented change at the top of the house, that still we see a lot of that in terms of adoption, in terms of solutions that are AI based and deployed and do actually work fine, but there is trust issues in terms of decisions that AI is making, not making, but you're also seeing all sorts of behaviors around driving adoption down for the exact fact that you mentioned. And one thing that I was very clear, and especially in my work and as we have taken some of this bigger problem statements with our clients and my clients, is you cannot treat people and talent and change as a byproduct. Once you have the strategy to execution and then you start managing change, it needs to be as you define your overall ambition around AI and recognizing the fact that 50% of jobs are going to change and think that expectation with the organizational leaders for them to trickle them down into their individual teams is absolutely critical in my mind.
Olivier Lafontaine:
You mentioned earlier that distribution is a key challenge that executives are targeting or attacking, I will say. So how do you see advisors and agents, how do their roles evolve over time and in an AI driven era? What's the real opportunity there in terms of growing the business as you mentioned earlier?
Abhishek Bakre:
For most of the presidents or head of distribution functions, if you will, that is the biggest challenge. And it's a little bit interesting because on one point you have a large protection gap in the US market when you talk about life and annuity in general. So there's clearly demand and the market's there, but at the same time, they're also struggling with the aging agent workforce and retirement and making the next batch of agents and distributors productive to sell life and annuity products because as you know, life insurance has to be sold, it's not bought, which then leads them to the question around how do I quickly recruit? How do I recruit in the first place, but how do I quickly get them up to speed? And I was talking to head of distribution with my client and the metric use, it takes anywhere from one to two years before a particular life and annuity agent that is onboarded to sell a carrier policies becomes actually productive and knowledgeable of their product set, if you will.
So in that sort of dual environment where you have the market opportunity clearly, but you're not able to hire fast enough and make them productive enough, the objective for most of these precedents is shifting from not just making them productive, but as AI comes along, throwing a third dimension into the mix, how do you keep advisors relevant in a world where advice itself is getting commoditized and routine advice is going to be simply digitized in some shape or form, whether AI and non-AI or other tools, how do you elevate the role of the advisor in that dynamic is challenging. And the focus for most carriers is not only to help them remove all of the administrative burdens around multiple underwriting and servicing and things of those nature, but also how do you drive trusted relationships between the agents and the prospects, if you will. And when you ask me what does the role of agent becomes, it really continues to become that trusted advisor to their customers when you have new prospects coming in, a lot more informed around life and annuity, how do you advise?
And advice becomes a product in many ways. And that role fundamentally then shifts to driving point life and annuity sales to becoming that long-term advise and an architect from a financial wellness standpoint with a lot of these customers, if you will. And if I look at the Inforce book, the lot of these carriers, that in my mind still continues to be an untapped growth opportunity to drive additional life as well as annuity products from an individual standpoint. Rounding that up, as I look at AI in this sort of distribution dynamics, AI should not only identify the products that the carrier wants to recommend, which is what most carriers are doing, but also how do you anticipate those changing customer needs over a period of time? And the future of distribution then really is a combination of human trust with intelligence from a carrier standpoint, and how do you bring that together to drive the most value to the customers, if you
Olivier Lafontaine:
So because as intelligent as AI can be, and it can certainly fake or it looks like it has emotional intelligence and it sounds like it cares, but I think other than a few exceptions where you could almost put that in the mental health category where people generally don't build a relationship with AI, they don't emotionally trust AI. And so that's important for the insurance products because it's there to protect, you're not going to be there when you need it, so you have to trust it. It's a very sensitive and emotional product. So I think that is probably where you want to focus your agent and your advisors. And this is where in the future, if the AI can do a bit of their legwork, probably is the right mix. So it's very exciting from that perspective. You mentioned the Inforce book.
Can you expand a little bit on that? What do you see as the opportunity for the Inforce book and how can this be exploited to get or used to get additional growth?
Abhishek Bakre:
Yes, several different ways, if you will. One is just driving better persistency within the customer base through meaningful engagement, but also through better experiences, if you will. One key aspect of growth lever in that InforceBook is also how to prevent lapses in the first place. And that is where some of the analytics and data signals and things like that come into picture as we sit on this AI agenda, how do you drive that lapses down from a year-on-year standpoint? And then lastly, what we've talked about as an industry for a long, long, long time is how do you drive cross? Someone having a term life, how are you going to convert that policy into more of a permanent life customer? And if it's a permanent life customer, how do you then supplement it with annuities or LTC or IDI type riders really focused on that holistic financial wellness?
Olivier Lafontaine:
You mentioned in our preparation, you're leading an enterprise-wide, I think that you're still doing that today. We're not going to name the customer, but the client, but an enterprise-wide quote to claim project or quote to claim redesign project built around agentic AI. Can you talk about that a little bit and how that's going?
Abhishek Bakre:
I mean, it's taking a lot of grit because it's not something that carriers are thinking about or approaching in the same shape or form. You see a lot of carriers, and I mentioned this probably earlier, taking use cases or point solutions type approach within a single domain. But when you truly look at code through claims, you're able to more effectively not just design the operating model, but also design the right set of technology and agentic and data architecture that supports that and do that in a manner, as we learned, that is economically feasible because this is expensive. And if you don't design it right, then your cost to value curve doesn't add up and end of the day, then you go and blame AI. And to me, it's designing that right at the get-go, both from a business as well as technology standpoint, really taking a hard look at people and talent at the get-go in terms of how roles are going to shift, which roles are going to stay, what are going to evolve, and setting that expectation as an organization.
Olivier Lafontaine:
So many carriers have modernized individual systems, but still struggled to establish consistent understanding of customers, policies, advisors, and claims across the enterprise. So how do you feel like data has been a limiting factor for enterprise to make good use of AI? And is there a role for a business architecture in solving this problem?
Abhishek Bakre:
I think in a very profound way, if you will. And what you hear in the market largely around data quality being a blocker for enterprise AI at scale, I see more than data quality, it's really inconsistent meaning from a business standpoint. And that is where your business architecture point comes at the forefront because the fragmented landscape that life and annuity carriers from systems and a data standpoint operate within drive the point that the 10 systems cannot be correct and at the same time still disagree on who the customer is because they're all defined very differently in each of these systems and things like that. And what you need from an AI and an effective AI standpoint is really more trusted business context and metadata, not simply more data. And the business architecture, as you mentioned Olivier, defines what that enterprise is and what is the semantic layer that enables at the enterprise level AI to reason consistently as it interfaces with employees, but also customers and agents and things like that.
And data tells AI what has happened, and Symantec really tells AI what it really means. So the popular belief, and my belief is a little bit contrary to that, fix the data first and then you can do AI. I say use AI to fix your data. Don't wait for data to clean up before you can start AI. It's vice versa because if your business and if you're depending on whichever life and annuity career you talk about, if your 100 DR business runs on data as it is today, AI can run the business too. So really the point that I want to take from that is really don't wait for data problem and quality as cited by most folks. Really use AI to clean up your data, if you will.
Olivier Lafontaine:
Good advice. And so you mentioned semantic, the semantic layer. Can you explain that for the non-technical person? What does that mean exactly for a business executive?
Abhishek Bakre:
Sure. It's having a common understanding of, like I mentioned earlier, how do you enable the AI to reason consistently across your business architecture as well as technology architecture? Think of it as phone book or think of it as metadata, think of it as whatever supporting semantic that you need and context that you need around your current data set that exists within the organization. And how do you give agent the right set of semantic and contextual knowledge to not just understand but act on that data, which is already fragmented.
Olivier Lafontaine:
And I guess it is a structured set of documents as opposed to just there's a popular belief that you can just throw everything at the AI and somehow think the truth will come out of that. But is it fair to say that the semantic layer is an organized way to give context to the AI machine, right?
Abhishek Bakre:
Yeah, because again, there's a lot of talk around this subject, just throw all the documents and expect AI to reason perfectly while you can still do that and AI will still reason in a great shape, but that semantic and then that contextual understanding layer really drives the value aspect that you're looking to drive versus expecting AI to do magic with just a bunch of documents. It still does, don't get me wrong, but not to the tune of what you want it to do.
Olivier Lafontaine:
Makes sense. And if we move on to thinking a little bit about the future, which may not be actually the future, maybe some of this is happening already, but what do you feel are some of the business processes or business areas that will have the biggest AI use cases? Is it underwriting, claims, servicing, general policy administration, IT? What do you feel is going to be the best use cases for life insurance companies?
Abhishek Bakre:
Three that come to my mind immediately, three big strategic bets for life and annuity leaders, and then two that will support how those bets realize and whether they work or do not work. One is, talked about this earlier, really reinventing your distribution and agency around AI-enabled advisor intelligence beyond productivity the way I described it, but how do you make advisors and keep them relevant as trusted advisors? So the investment in a supporting bucket of work. Second bucket of work, and we can talk about that a little bit after this question is how do you truly re-imagine service as a competitive differentiate? And on that aspect, there is again two ways to look at it. Service in a closed block is very different than service in an inforce block. Closed block, you maintain parity at the industry levels, but at the in-force book, we can do a lot more in terms of leveraging that service as a competition.
So that's number two. Number three is really the more core and fundamental aspect that we talked about, how do you build trusted enterprise intelligence through the business architecture that you mentioned earlier, through the semantic models and through governed and trusted AI on top of it? I think those three, if I was CEO of a life and annuity carrier, I think those are the three investments that I would pay a lot of attention to. And underpinning that, I think there are two things that are critical to drive success around those big bets, which is how do you simplify products? We talked about decisions from an AI standpoint, but also the supporting operating model before scaling AI, and number two, really focusing on those enterprise capabilities, end-to-end re-imagination, laying the end-to-end fabric rather than a set of disconnected initiatives and use cases, which unfortunately has been more of a norm these days.
Olivier Lafontaine:
Yeah, that makes a lot of sense. And I like the idea, I guess the idea of using AI and services, I guess it accelerates or I guess it affects all customers and it supports what you mentioned earlier around using the Inforce book as a way to drive growth and improve financial outcomes and to persistency and cross-sell and upsell. So I think that has real potential. I agree with that because it's one area that is expensive and unfortunately, well, not unfortunately, but the nature of life insurance is that it lasts a long time. So when you price the policies in the first place and the systems you're on at the beginning of the policy are probably not the same that will be around 25, 30 years later. So the servicing cost is always a big concern and that's why it suffers sometimes a little bit. But with AI, perhaps two things can maybe true at the same time.
We can have good service and it still doesn't cost more. So that's exciting in that way, right?
Abhishek Bakre:
Yeah. And the key is not AI everywhere, AI where it makes sense. Agentic even more so where it makes even more sense from a decision making and autonomy standpoint. But if you look at service, to your point, and glad you brought that up, it's really evolving or has been evolving or carriers have been trying to evolve it for a while to move from that transaction processing set into what I call as the decision oriented engagement as you look at AI and things like that. And the objective is really not answering questions from a customer standpoint, it's preventing them in the first place from operating, especially the simple intents and people servicing requests and things like that. And if you do that correctly, then it becomes a source of your growth that we talked about earlier, a source of loyalty and ultimately source of trust with your customers and agents.
And you would agree with me that not every customer segment or cohort, if you will, deserve the same service model, the segmentation around the customers to really drive the economics around it. And interested in a service organization in my mind really proactively orchestrates those outcomes versus been reactive.
Olivier Lafontaine:
Wrapping up here, we've been talking for some time and I think this is all great insights and great input and though-provoking ideas. If we look ahead and you speaking to leaders and talking to business executives, what are the things that will make some leaders successful and some less successful in the next five years? What are some of the things that people have to do to take advantage of this new wave of technology and what it holds and take advantage of the promises and not fall victim to the pitfalls?
Abhishek Bakre:
What's going to differentiate in the coming years is that we have four or five things in my mind. One being the competitive advantage is going to shift from products, it's going to shift from modernized core to more of an enterprise intelligence. And that's going to drive the competitive advantage for a lot of these carriers because products can drive all sorts of product innovation, drive all sorts of system modernization, but a lot of that starts to get commoditized and the enterprise intelligence and the inherited an intelligence that most of our clients have internally, how do you harvest that is going to be key. And AI by nature of it won't really differentiate these insurance companies. Again, going back to my key point around how do they redesign the business around AI and that will differentiate them. And the leaders will really start organizing around decisions as AI becomes even more embedded in our day-to-day and in the business world instead of corporate function and domains and things like that.
And as all of these come together, the human expertise, both with the employees and tenure knowledge that a lot of these employees have becomes even more critical and valuable as you use AI to automate routine cognitive but also non-cognitive work. And there's a difference in, like I specifically said, cognitive and non-cognitive. There's some thinking pieces also AI is going to handle really well. And as our clients and insurers that learn, design, adapt fastest to this amazing technology change, I think will continue to outperform the market, if you will.
Olivier Lafontaine:
Great advice. Thanks. And I think we'll leave it at that for today. That was all a lot of very insightful thoughts. Thank you for that. Haby's years of experience across core modernization, digital, and now AI is a good reminder that this isn't really a new problem. It's the same organizational challenge the industry has faced for decades, just wearing a new coat. What stood out to me most with his point on data. The popular belief is that you have to have a clean data before AI is added to the loop for help. Abhi flips that. Sometimes AI is the thing that cleans up your data. As more carriers have moved past scattered use cases toward real enterprise intelligence, Abhi's experience shows that the leaders who win be the ones with the best technology. They'll be the ones who redesign the decisions and operating model around it.
Thank you so much for listening.
