Get a demo
Select the type of demo you'd like to experience with Indico Data.
Select the type of demo you'd like to experience with Indico Data.
Watch Indico Data CEO Tom Wilde step in as co-host alongside Michelle Gouveia, VP at Sandbox Insurtech Ventures, in season 2 episode 5 of Unstructured Unlocked with Farooq Sheikh, Unqork.
Michelle Gouveia:
Hey everyone. Welcome to another episode of Unstructured Unlocked, Michelle Govea and Co-host Tom Wild. And we are joined today by Farouk Sheik, the global head of Banking, insurance, and healthcare at Uncork. Faruk, welcome to the podcast.
Farooq Sheikh:
My pleasure, my pleasure being here.
Michelle Gouveia:
Looking forward to the conversation. Just to get us started, would you mind sharing a little bit about your background and what it is you do specifically in your current role?
Farooq Sheikh:
Happy to do so. So I grew up as an actuary and started learning about insurance and so forth. Spent a lot of time in a management consulting firm, Oliver Wyman. I was a partner there and now I lead insurance, banking and healthcare businesses at anco NOCO platform across the industries.
Michelle Gouveia:
Excellent. Well, I’m sure you see a lot of interesting processes and have a lot of conversations with folks making key decisions about workflow initiatives and things of that nature. I think the first question I have, because your title so aptly puts all three in there. What are some of the big similarities or differences that you’re seeing in the market today between healthcare, the p and c insurance and then banking as it relates to key initiatives that the organizations are trying to push forward?
Farooq Sheikh:
That’s a very, very good question. And you’re right, there’s lots of similarities across these industries. All of them have made some progress in terms of digitization of process and increased amount of automation, but all of them have a long way to go and you can look at it across the waterfront, so to speak, from commercial banking and retail banking, institutional banking, even today for some of the largest companies, the onboarding of a new customer, the underwriting credit, underwriting in this particular case for one or multiple products, the ongoing servicing elements are mostly manual. The same applies in the wealth and asset management space. The same applies in the insurance space and the life and PNC side. And of course healthcare. There’s quite a significant amount of manual processes, like pre-authorization is one example where you have this interface between the payers and the providers on adjudication of service and care, and all of these things have been there for a while.
So this is not new, but what’s happening now is that there’s an increased amount of pressure both internally and externally. So externally from a customer perspective, there’s an expectation that some of these processes are more automated. They don’t take as long, for example, to close a mortgage, doesn’t take as long to get a credit approval on a mortgage, doesn’t take as long. The same expectations are changing on the business side in terms of onboarding for a new customer in the banking space or in the insurance space. So the expectation externally is changing pretty drastically based on some of the other expectations driven by other industries. And the expectation internally, I think is the big one. Internally, the underwriters want a better experience internally within a bank or an insurance company. There is a finite amount of talented folks who don’t want to spend their time navigating operational or manual hurdles.
That’s kind of a big change now. And there’s pretty much the winners here across industries are the ones that can automate and digitize the process with humans in the center. So this is not a conversation about removing that human expertise, it’s more about digitizing the process. And this applies to every part of the value chain like credit underwriting and banks servicing, the claims adjudication on the healthcare side, the PNC underwriting side, the claims side, every piece of the value chain, the employees and the colleagues are looking for a better, more automated process. And that’s kind of the big, big challenge.
Tom Wilde:
Talk a little bit about the founding thesis that you had with Uncork in terms of specifically the way you assumed insurance and banking and healthcare used to buy and implement technology and the world you saw after Uncork. So curious, what drove the founding principles of Uncork?
Farooq Sheikh:
Very, very good question. And as you know, most of the folks at Uncork come from industry. So we lived and spent a lot of time navigating some of the challenges. So there are two main principles at play. One is around tech debt. So that is a lot of the times. The idea is that when you have delivered something in production, you’re done. However, in the course of, in the context of an enterprise, that’s the beginning of the journey, like the enhancements, the tech debt, the the vulnerabilities that all of that starts then, and that requires a life cycle of its own and it’s a 10, 20, 30 year journey. So how can you simplify, and there’s a lot of very good artifacts on this, but one of the artifacts that we really thought about were how do you increase the rate of change? So if the company, every time they put something in production now has to think and worry about all these things that they need to do for that one app, they’re not able to move faster onto the next or the next layer of automation and digitization.
So can we take that burden? Can we augment that burden from a codeless capability where the security, the, the vulnerability scanning, all of those things are not the remit of the enterprise, more the remit of the provider in this case. So that’s one big theme. And the second big theme was that historically the buying pattern in enterprise was you pick your vendors, you pick your partners, whoever they are, and you make them work together. So inside your enterprise you figure out how they work together and what the architecture is and what have you. And regardless of how big your vendor is or how standard your vendor partners are, you are figuring that out. It’s kind of an enterprise’s job to figure out how vendor A, B, C, D work together in service of their goal. And I think we’ve seen, yeah, that has changed now or is changing now where it’s more an ecosystem play where yourself and us, we figure out how we work together and that makes the life of our enterprise customers easier and it allows for a better, seamless kind of experience for the enterprise customers when they are trying to pull two or more enterprise players together.
Michelle Gouveia:
That obviously puts a little bit more pressure on the vendor to have some ecosystem partners in place or proof points that they’ve partnered with others in the past, people that those enterprises are already working with or they could or that are similar to who they’re working with now. How should startups be thinking about building those partnerships and how to sell that into the enterprise as part of their own process?
Farooq Sheikh:
And it’s actually, I’ll start with the latter part of your questions and come back. So especially with insurance and banking, frankly, the ecosystem plays are very well understood. Insurance, the interface between a broker to a carrier to a reinsurer, to a claim adjuster, that’s an ecosystem. And they’re used to working together on individual deals as well as multiple deals. So this whole interface of there is an ecosystem and you’re a member of that ecosystem. So you need to work with many others in your space and be easy to work with is very much part of the ethos of insurance. And by the way, the same is true for banking. When you think about syndicated loans or commercial lending, you have multiple lenders involved and all that. So that whole ecosystem play. Any you have warehousing in mortgages that might be a different, somebody’s servicing the mortgages as opposed to somebody’s originating the mortgages. So there’s a whole ecosystem in the businesses that we serve and that’s an expectation. So now when it comes to the vendor side, it’s the same idea, which is you are also, and technology is a very natural place for an ecosystem because there’s no company, and this used to be the fallacy where one company would try to own everything and that kind of created the mess we are in frankly, in terms of monolithic systems and so forth. You’re
Michelle Gouveia:
Very polite in all
Farooq Sheikh:
Well. So if you recognize that there’s no one company that can do every bit of it and you need your partners as much as they need you, and it’s all in the service of your customers, that makes your life simple. And then I think your question specifically was around how do you think about partnerships? How do you think about pulling these things together? And I think honestly, the simplest way to think about it is put customer in the middle. As long as every conversation is thinking about and starting from an ending at how is this better for the customer, it simplifies itself. The place where these things fall apart is when you have some objective, which is not customer driven around ownership or like, Hey, I want to own this thing, and that has nothing to do with the right way of doing things or the customer’s preferences or whatever. Just like an bias in the system. That’s where these things start unraveling, and that’s not the right way to think about it.
Tom Wilde:
Talk a bit about the tension and or balance between the customers. We think about this a lot at Inco, the customer’s desire for package solutions out of the box versus the customization and flexibility that comes from that. You guys chose a low-code approach to solve some of these challenges, but where do you see customer thinking around out of the box versus custom? And is that changing at all? How do you balance that at unor to meet their needs, but also in some ways sort of protect them from themselves a bit? Right?
Farooq Sheikh:
So Tom, is it fair to assume that the typical demand of out of the box fully configured to what I need isn’t realistic? Is that what you’re hinting
Tom Wilde:
At? As a vendor provider, I spend a lot of time explaining why in concept it’s brilliant and in practice it just isn’t feasible.
Farooq Sheikh:
It’s funny, it’s a little funny frankly that you continue to get that request, which is I have a unique and bespoke process, however, what I would need is something that’s out of the box and fully configured to the process that I have in mind for my future state. So it’s a fun conversation always. I think there’s a little bit of education there, just being frank about it, there’s a little bit of an education there. I do think that the industry has made very good progress in understanding what type of technology flexibility is available to you in certain contexts, like this kind of no-code layer capability and the traction there is a function of understanding that if you need something more bespoke and flexible from a UI and our workflow and presentation and data model perspective, this needs to live in a system that’s not, for example, a core policy admin system where the degrees of freedom might be lower, right?
Similarly, the ingestion frameworks and the ability to target the emerging LLMs and so forth that are coming out to get a better outcome of ingestion, they understand that that needs to be done in something like NCO and so forth. And that’s not going to putting an LLM and a policy statement systems not going to get you the right outcome. So this whole bifurcation and understanding of what technology should sit where to enable, what flexibility is honestly getting better and more and more understood. There’s a journey here we need to get through. The question that you were posing though is how do you think about this tension between out of the box or prebuilt versus something that’s more configured? And I think the short answer on that is that at the end of the day, it is about business outcomes. So as long as you can get to a point where the business outcomes are measured in days and weeks, it doesn’t matter how you get there, frankly, if it’s not fully exactly what the customer wants, as long as there’s flexibility to modify it and configure it, that’s a fine outcome.
I think one of the mistakes that folks make is that the inner attempt to become more out of the box, they harden the architecture of the product and make it less flexible, and that kind of beats the purpose of this business outcome that you’re trying to achieve. So that balance, maintaining certain amount of flexibility in your product architecture so that you can get to the business outcome in days and weeks is probably the best way to think about it. I haven’t found a way to preempt it. That is you’re never going to have everything that everybody wants, but as long as there’s flexibility in the system to get there fast enough, I think that kind of serves the purpose
Tom Wilde:
And that flexibility comes with important architectural thinking. And I know you guys have put a lot of thought into this around security and governance as well.
Farooq Sheikh:
Yeah, yeah. The security and governance pieces paramount. If you think about enterprise SaaS, the starting point and the ending point is security and governance and the shared security model between cloud, the application layer and the enterprise. That is a very significant, and I do think that platforms have done a good job in increasing the security posture, for example, that a company may not have all the bells and whistles in place because they’re building individual applications by code and they need to go through. But when you get into a platform structure, of course you’re doing that on behalf of so many more of your customers and so forth, so you’re seeing things. So that kind of increases the security posture. Plus of course, reliance on the cloud security always helps.
Michelle Gouveia:
Brooke, I wanted to go back because you said LLM, which is the magic trigger word to convert us over to an AI and automation conversation. Specifically, we’re going to talk about ai. No, we never do
Farooq Sheikh:
20 minutes in. Nobody has said ai. I was worried.
Michelle Gouveia:
We’re getting there. We’re getting there. I could ask you just generally speaking how you think about the role of AI across the various sectors that you’re supporting, but one of the things I want to also call back to is the discussion on ecosystems and how do you think AI and automation will impact the relationships that happen between, well, I’ll just call the value chain as you outlined it between brokers, carriers, reinsurers, and how they all interact together.
Farooq Sheikh:
This is a very good area of discussion. I would love to get your input as well, but I can share how I think about it. So I think a couple of comments. One, I think is just being realistic. There’s a little bit of too much hype for the short-term impact of tech, and I think we underestimate the long-term impact of ai. So I think both of those things are happening where we are not really overestimating the and underestimating the long run. But the other piece of it is AI has been around for a while, whether it’s machine learning, AI or generative AI is coming out. And I think what’s happening now and then we are still at the early stages of it, is that you have multiple tools in your tool belt that you have multiple models, open source or otherwise, you have multiple kind of inference capabilities.
You have the kind of prompt engineering as a capability emerging to become much more significant and so forth. So you’re seeing all these tools start to become more significant, whether it’s machine learning, the good old machine learning, the ai, degenerative ai, and within generative ai, the types of models and the open versus closed side and the training and entrance pieces of it and prompt engineering on top of it. So now if you think with that context where we are seeing a lot of excitement, so ingestion of course is a big one because gen AI and Netherlands allow you to do a better job with extraction, but also to include the contextual awareness of the document as opposed to just the optical recognition of it. And that is like a slam dunk. The amount of activity in extraction and IDP intelligent document processing and the enhancements we are seeing there due to LLM utilization is one of the most significant applications of gen ai.
Just being simple about it, the ability to understand what the data means, what does it mean in the context of the risk that’s being evaluated, whether it’s insurance, banking, healthcare or otherwise, that is very, very, very significant in terms of the improvements. There are other areas, just to complete the thought, there are other areas of gen AI as well in terms of developer productivity. That’s a big area in terms of chat functionalities and summarization use cases in terms of and so forth. But I think the ingestion area is truly benefiting a lot from the development of genai. I think we are yet to see some of the more comprehensive use cases of genai in the application layer. So you’re seeing that in the marketing side, you’re seeing that in the content generation side when the application layer because they struggle with the same issues like the legacy data silos and legacy system issues. So that is the piece that will happen over time and we underestimate the benefit of that, but as better data sources come together and more centralized client data platforms come together, but all of these other things like marketing, summarization, content, generation, IDP, developer productivity, there’s a lot of value
Tom Wilde:
There. We talked about packaged versus custom. We talked about the rise of generative AI and this sort of copilot metaphor. Let’s talk a little more specific now about the workbench, right, the underwriting workbench, which I think is one of those things that everyone knows what they mean, but no one can quite define what it is. So I know that’s been a hot area of interest and use case for uncork. How do you think about the bookends of the workbench and are the requirements and desired outcomes, are they evolving or how much in focus is this idea of a workbench right now?
Farooq Sheikh:
It’s a big area of focus, and I’ll explain a little bit, right? So what we are seeing is that there are two kind of main elements here. There’s an element of, to my comments earlier around the joy of underwriting. You have underwriters that are struggling with manual data entry, silo systems moving things around, and they’re kind of saying, guys, this is not a good use of time. That’s one big area. And the second big area is that if you look at any carrier, if you look at every carrier, they only can respond to some percentages of the submissions they get. So just from the leaving business on the table, because they cannot process everything. So you have the people who are processing stuff saying it’s too complicated and difficult to process. And from a business perspective, more than half of the submissions are not being processed to begin with.
So these two forces together are basically saying that there’s a better way, and there is a third element as well, which is important, which is better risk assessment. But in a lot of companies, the underwriter, they’re very talented. So even they’re kind of overcoming the manual nature of the process to do a good job in risk assessment and are taking their time doing so there’s an argument to be made around better most standardized risk assessment. But that I don’t think is the primary argument. The primary argument is the joy of underwriting from an underwriter’s perspective and the ability for the carrier to process more business or maintain the share of its business, especially as others transform the distributors and brokers transform as the insurers transform, everybody has to keep pace. So that I think is the biggest area, and that’s why this concept of underwriting and underwriting work pension is coming up now it’s defined broadly by the way.
So it typically goes submission to buying and beyond as the way we describe it. So you have the starting pieces of submission, ingestion, extraction, enrichment, triage, clearance, et cetera, and running it all the way from rate 10 quote and bind and then endorsements, renewals and so forth. It’s a full value chain process that breaks between simpler lines in terms of an increased STP process and complex clients in terms of more digitized risk assessment and underwriting. But there isn’t a conversation where the underwriting teams or the operational teams go, no, no, we don’t need a work bench to facilitate and expedite the process. We haven’t seen one of those yet. It’s mostly, hey, we are looking to build it. We’re looking to build it for one line or few lines or 10 lines or all lines. We’re looking to build it something that we can do in one market or multiple markets, but it’s not a question of, so it’s never a question of we don’t need something like this to facilitate an work.
Tom Wilde:
There’s some tension between the notion of automation though and the fact that underwriters and actuaries, I mean they’re very experienced people, judgment’s a key part of what they do. And so finding that right balance between giving them a bionic arm versus full automation, I think there’s a weariness and attention around that with these technologies.
Farooq Sheikh:
It’s a very fair comment and just to give you a little bit of my perspective on that is that a little bit of this also, it used to be a big topic like five years ago and now it’s becoming another big topic with the event of ai. So this topic typically comes from one of two places, but typically comes from a fear of like, Hey, are we going to automate the human event? And that five years ago there was a debate. It’s no longer a debate because it’s clear that they’re trying to digitize the process around the human with human in the middle. It has come up again with how significant gen AI is and there’s this fear of, Hey, is gen AI going to fill in the gaps? Whatever kind of fear you would perceive it to be. But I think increasingly people are realizing that you still need a human in the loop with all these kind elements of gen AI that need to be addressed, whether it’s training data, whether it’s security, whether it’s hallucinations and so forth. So you’re still in that same place where as companies get more comfortable using it, you’ll end up in the same place of you still need to automate, digitize the process with the human in the middle,
Michelle Gouveia:
Keeping on this topic of that large workflow where the underwriting work bunch takes place and relating it back to all of the hype to use your word proof and the promises of ai. Whenever I talk to companies that are saying we have the AI solution to do it all, we will be the decision making, the workflow solution that does all of it, not just this one part really well, that tends to make me nervous to hear. I think it makes companies that are partnering or looking to work with these companies a little bit nervous, but what do you think about companies that are promising to be everything to whether it’s an underwriting workflow, a claims workflow, even with keeping the human in the loop?
Farooq Sheikh:
Yeah, a little bit of this is the comment I would made earlier in terms of the ecosystem and the value of the ecosystem because you end up eventually, these technologies are slightly different in terms of the flexibility they provide and Nick the ingestion technology and the sophistication it needs in the ability to leverage the large range of and types of AI capabilities versus what a workflow capability might be that needs you to facilitate fragmentation of experience and business rules and logic and integrations and all that. So there kind of two different things to begin with. And if you go back to my comments on the ecosystem side, historically, the way we got into the mess as I described it is when few want to own everything, not necessarily because the technology is a natural kind of way to consolidate a technology conversation or architecture conversation, but more from a place of we want to own this part of the value chain or that part of the value chain, and that never ends up well because it’s not a natural way to facilitate some of the flexibility that we talked about. So I’m with you there. I think the monolithic that’s put everything in one box, the box is optimized for A or B, but let’s hope that it does both Well, it’s a complicated place to be,
Michelle Gouveia:
And I think while
Tom Wilde:
Most insurance companies, banks don’t aspire to be software companies, they sometimes I think veer into that direction, but most I think don’t really aspire to be that they do believe that their ability to assemble both the data and the application experience in a unique way can create competitive advantage. And that does feel viable and eg legitimate. I mean obviously you can compete on many different fronts, but to your point, I think if it is just a big monotech software program that everyone in the industry can buy, then it’s no longer really going to provide competitive advantage. So there’s a little bit I think of that where they don’t want and they don’t want any one provider gaining enough market power that now they have a problem. So I think it continues to foster a lot of innovation and the need to be able to co-exist across many different applications and APIs because the insurance companies and banks like to do it specific ways.
Michelle Gouveia:
So you’ve talked a little bit about the use cases that we’re hearing the most about or seem to be the best fits right now for automation. How do you see that evolving or how does uncork think about that evolving? How does that inform some strategy about product development or where the market is heading three to five years from now?
Farooq Sheikh:
I think there are two parts to it, right? The first one honestly is that I think in the next three to five years, we just need to do a better job of addressing the legacy technology and silo data problem. And it doesn’t mean that everybody’s going to have brand new technology for everything, I don’t think. And that historically was the thought process, which is like if system A doesn’t work very well or if it’s old or difficult to maintain, replace it with system B. And I think what’s happening now is that people are saying, fine, we have 30 or 40 systems that are old, but can we push the data out? Can we get the data layer to be more modern? Can we get the integration hubs to be more modern? Can we get the digital experience and calculation and rules capability to be more modern?
So it’s like a top to bottom digitization where you’ll end up with the two speed technology capability where you have some systems and capability which are slow and that’s fine, but you have all the key capabilities that are in the more modern stack. So that’s going to happen. And more importantly, that will free up the data because all of the AI in the world doesn’t really work if you don’t have the ability to bring the data into the mix. So that’s one piece of it. And I think the second piece of it is the outside in, because I think what’s happening now is that the ability to extract data in a systematic way, enrich it, bring in the right data vendors and orchestrate the right enrichment and internal processes, that gives a lot of power. So you can see that the internal data gets freed up or gets more accessible.
The external processes get more automated and they meet in the middle, and that’s where the real competition’s going to be, which is who can get there faster, better in a more systematic way to think about what risks do we want? That’s the first objective. How do we get them at the point of underwriting at the kind of working with the brokers when distribution, and then how do we manage our portfolio as the overall risk environment changes? That’s kind of where the objective is, where you can free up your data internally to bring it and then digitize your process of getting external data and external submissions and digitize that.
Tom Wilde:
So the data’s a great one. Are there other things that gen AI is imposing on these companies? So what do they have to do differently? I think clearly the data preparation access to data, it is still a data in decision out sort of a construct, but maybe on the sort of process workflow change management, what kind of impositions does AI suddenly bring to these companies that they’re going to have to sort of rethink?
Farooq Sheikh:
It’s very, even today, there are markets and businesses that are getting more complicated because of the imposition of external data engineer, small business, think about small business. So historically small business, there’s a number of players in that market and they operate in what I would call packaged, standardized packaged products. But even now, there’s been enough MGAs, enough underwriters that have gone after slivers of small businesses, Hey, we really focus on construction, we really focus on dentists and whatever program businesses and these businesses have scaled well, what does that mean? What that means is that the person or the set of businesses that we’re going after the generic small business environment now have lesser degrees of freedom because enough of these niche programs have come together. Now you add gen AI to it that just puts that on a steroid. You can see already California wildfire, the condo associations specific types of programs are getting constructed that can go around and focus on specific niches which reduce the degrees of freedom and the amount of business that’s available to a more generic small business or mid-market focused writer.
So you’re seeing long way of saying you’re seeing the definition of competition change where you had a small business writer competing with a small business writer. Now that’s not the case. You have a small business writer competing with all these specialist MGAs that is competing with somebody who’s not yet come into the market yet, but can do so very quickly because they have the ability to access these risks, underwrite them using the latest and greatest technology and maybe less reliant on the competitor advantage, which is underwriters. So there’s a machinery here which is changing the dimension of competition in every market, and that’s true for life insurance, that’s true for commercial lending, that’s true for healthcare, where more specialized offers and programs can be constructed more rapidly. Now the tech stack isn’t that complicated, especially if you don’t have to work with legacy, you can be up and running pretty quickly and these things take time to scale. So there’s not that you start something tomorrow, but there are examples where enough MGAs have scaled well in this space. So you’re seeing that kind of veteran play out.
Michelle Gouveia:
There’s the competition question that we were just addressing within each industry. So we just talked a lot about the MGAs and how you’re kind of breaking up that sector, but within the broader industry sectors where based on what you predicted for the next three to five years, is it they all have the regulatory hurdles, they all have their association to being slow moving industries. Does insurance versus banking versus healthcare, which one adopts faster, gets through the compliance and the regulatory hurdles and the governance requirements? Is there one that maybe outpaces the others that becomes a model that the others follow? Or are they all kind of moving along at a similar pacing?
Farooq Sheikh:
It’s a very good question, and I laugh because I don’t know if one is faster than the other. I can tell you enough of them are slower than others, but it’s a
Michelle Gouveia:
Tight race, right? It’s a tight long race.
Farooq Sheikh:
But there are, just to be fair, there are very real considerations that need to be managed, especially from a regulatory and compliance perspective. Like gen AI for example, doesn’t allow for traceability. So if you’re doing good credit underwriting, you can get to an answer. You can’t explain how you got to that answer. So that’s a big part of Gen AI that needs to be addressed and they’re kind of companies that are in that space trying to augment the benefits of gen ai but maintain traceability. But that’s I think where a lot of the adoption conversations are sitting right now that AI can accelerate, AI can digitize, but AI can’t be used for decision making purposes at the moment because of the limitation on traceability. And as soon as you crack that kind of situation where you can augment the various tools in your tool belt to get to the benefits of AI plus the traceability component, that is a big, big revolution and that will apply across the board. In terms of which industry, I think the problems are the same, like data and legacy technology, data silos and legacy technology, the desires are very similar. There are certain pieces which are simpler and have invested more. So anything with higher volumes, retail banking is the best example. They’ve just been investing in their digital infrastructure a lot longer because they had to, right? The demise of branches and retail health is the same, right?
Tom Wilde:
And the rise of mobile, right? Mobile’s really
Farooq Sheikh:
Exactly accelerate. Exactly. Very rarely are you working through non-mobile and non-digital channels with your retail health and retail banking situations. So those places are
Tom Wilde:
As well, right?
Farooq Sheikh:
Yeah, exactly. Retail insurance, like auto home, et cetera. But those places are the ones where we are going to see more progress sooner because they just have more of a foundational readiness and then all the institutional stuff or all the stuff that doesn’t have the most amount of volume will take a little bit longer just because they have to get up to parity on the foundational investments before they can proceed beyond,
Tom Wilde:
I sometimes refer to as this Tower of Babel problem, right? Whether it’s in banking, insurance or healthcare on the commercial side where the channel’s very dispersed, you have a very long tailed channel. If you think about broker to carrier or mortgage broker to lender or physician provider to payer, there isn’t a moment where someone can gain enough sort of channel power to dictate how it’s going to work. So the market has to sort of figure out how to evolve and coordinate itself over time in this very, very diffused sort of landscape of players. And I think technology’s going to have to be the fabric that does that connectivity. Because today, I mean you see this I’m sure every day too, that commercial insurance, there’s two ubiquitous pieces of technology that are in every single account you find and it’s email and Microsoft Excel, right? Those are the two platforms of record that you’ll find everywhere.
I often joke that email is the ultimate API still in insurance because that’s how the brokers and the carriers do most of their data transfer is still through email and through spreadsheets. So certainly there’s big opportunities to drive transformation and gains in efficiency and for the carriers, they really haven’t yet. I think McKinsey did this study or Deloitte, I can’t remember which, that despite huge gains in productivity across almost every segment of the economy, you haven’t seen it insurance yet. The combined ratios haven’t moved that much in the last 10, 20 years. So there’s still more to capture there for sure.
Farooq Sheikh:
Yeah, I think there’s value there. You can see clearly and to your point around the expense ratios, you can see clearly the impact on profitability in the industry and how that can be improved if you are able to figure this out.
Tom Wilde:
Well, great. We really enjoyed the conversation. We’ll let you get back to my tie and the beach lounger there.
Farooq Sheikh:
I’ll just turn around and go back onto the
Tom Wilde:
Beach. Exactly, exactly. Take a dip. But it was great talking to you. Love the conversation and thanks for your time.
Farooq Sheikh:
Likewise, really appreciate then invite. I really appreciate the conversation and I look forward to many more.
Tom Wilde:
Great
Michelle Gouveia:
Thanks.
Farooq Sheikh:
Cheers. Take care guys.
Check out the full Unstructured Unlocked podcast on your favorite platform, including:
