In the Hot Seat with Margeaux Giles

Speaker A: Insurance Unplugged in the hot seat, where the complex world of insurance is laid bare. Hosted by Lisa Wardball, this podcast promises an unfiltered glimpse into the industry like never before. Each episode invites you to listen in on the candid conversations that usually happen behind closed boardroom doors. From deep dives with industry leaders and thought leaders to innovative discussions with minds shaping the future of insurance, we bring the most genuine talks directly to your ears. Our guests take the hot seat alongside me to explore the inner workings, challenges and triumphs of the insurance world. If you've ever wondered what goes on in the shadows of the insurance industry, from the boardroom banter to the behind the scenes strategies, this is your chance for a front row seat. Prepare for unguarded, enlightening and engaging discussions that cover every angle of insurance presented in a way that's both insightful and accessible.

Speaker B: Welcome to the conversation.

Speaker A: Welcome to Insurance Unplugged in the hot

Speaker B: seat with Lisa Wardlaw.

Speaker A: Welcome to today's episode of Insurance Unplugged, proudly sponsored by Iris Insurtech, your gateway to the future of insurance distribution. At Iris, we harness the power of generative AI to revolutionize data processing and decision making across the distribution spectrum. Our platform integrates Genai to provide not just insights, but actionable intelligence, configurable workflows and dynamic form generation, all underpinned by continuous data quality management. Discover how IRIS is pioneering smarter, more efficient operations in the insurance industry, paving the way for a new era of distribution excellence. Let's dive into how Genai is transforming the landscape of insurance distribution today on Insurance Unplugged.

Speaker B: Welcome to another episode of Insurance Unplugged. I'm your host, Lisa Wardlaw. And this week I like, feel like we need drumrolls. I have Margo Giles joining me, CEO and founder, Iris Insuretech. And I'm going to let her like, she's no stranger to the show, but I want her to kind of introduce herself once again. For those of you who may not have heard her intro or may not know her deep level journey, but really what I wanted to first lead in was saying like Margot and I are not here to do a highlight reel. She and I are really going to break down. I'll call the deep nuance of, uh, what it's actually like to build an operational system. Not for demos, not for, I'll call it Point Solutions, but what it's really like and I'll call it like her candor and survivorship through this is huge. So welcome to the hot seat, Margo I feel like you've been living in the hot seat going on this boundary.

Speaker C: I am the epitome of the hot seat, I think.

Speaker B: So if you don't mind introducing yourself, I think like if people have listened to us previously, they've heard your journey, but maybe just a high level intro and synopsis of what you set out to build so that they can understand for context the questions and the follow ups. I'm going to kind of put you through on this season finale.

Speaker C: Sure. Awesome. Margo Giles, founder and CEO of Iris InsurTech I have been in insurance for almost 20 years. So I did not start my journey on the tech side. I started in the seat. I have always lived on the distribution side of the house in PNC insurance. So, you know, writing mid to large market commercial, being a producer and then being an owner and then going through M and A and now going through M and A again. Um, always living in that world, you know, so running agencies and brokerages, dabbling in MGAs and MGUs. But yeah, I mean, an insurance girl through and through. And I kind of got thrown into an ops role as my career progressed and you know, being younger, uh, the younger side of a COO in insurance. I think I was 29 when I took over the first insurance agency. So very familiar with technology in my personal life. And then I would come to work and it was just such this huge disconnect between what I was able to do outside of insurance and what the software and the systems were allowing me to do inside my own business. And I felt the friction of that and the disparity between those two things was just way too much to ignore. And so I went on a very long journey of learning by trial and error and by fire how to, how to build tech, what it really means, how to do it internally in my own teams and then how to scale that out, you know, into a real SaaS company. And, and that whole journey has probably been about eight years. So yeah, I definitely am, um, battle worn at this point. Feel very, very confident that I can answer a lot of questions. You're probably going to ask me about how this process unfolds.

Speaker B: Yeah, yeah. And just to kind of like recap because I think people are like, well, like Iris sponsored it. But we like went through a journey. We heard so many voices like over this season and prior seasons. Right. And I just wanted to call it kind of like connect the dots for people. Right. So just recently we had Chloe Perry on last week. If it's not auditable, it's just a toy, you know, audit as though you're being subpoenaed, designed for that. We had Casey Kempton, who you sat on stage with her, Margo, last year at iti. And you know, and she really got underneath the fallacy of predict and prevent without backend design and Andrew Morse treating Gen AI as a teammate, not a bolt on the core theme. Underneath this was really real transformation. Like, authentic transformation means that you have to face these financial truths, system lags and infrastructure failures, not just adding AI to sales slides. Right? Like, you and I see that a lot. And, and to me, clearly, I know this because of my work with you. Iris sits underneath every one of these themes and how you really went beyond selling a vision. So, Margot, now I want to get to the. Like, right? Like, um, you and I had this conversation a lot, right? Like, most founders have to sell vision, but most founders stop there. Few survive the execution layer, which is why we see. I'll call that high velocity against that cliff that you and I started a couple of seasons ago talking about that 3 million ARR. Uh, Cliff, why we see so few people push into the execution layer and even fewer can build something that the industry, you know, isn't yet asking for, but needs. Right? Because like, a lot of people would say, we'll build for what's asked for. But it's like, yeah, uh, but, you know, again, we'll quote Henry T. Ford. They would have asked for a better horse, you know, and we hear that a lot, especially in the distribution.

Speaker C: They're asking for a better horse. I mean, realistically, living in the seat of sales and marketing. But that's what they're saying. But that's actually not what they mean. And it's very difficult to get underneath. To your point, as, what are you. What do you really need? You're describing a symptom, not the root cause. Uh, that's tough.

Speaker B: So how do you. So let's take that one because, you know, you and I talk about this all the time, and it's difficult because you've got the lenses, you've got the customer lens, you've got the VC lens, you've got your funding lens, you've got your product lens, you've got like, how do I market this? How do I commute? Like, the dimensionality of all the plates, you have to spin literally in parallel on. I'll call it quicksand, right? Which is your timeline and get to market. And you've got to have something that can be usable. So what. Let's step Back in your voice, what do you think the industry actually wants to change? Or how do you give them what they think they want? Is there a. But there's a reality. It's like, well, I still have to have a car that drives even though I might want a new radio in the car. Like how do you, how do you compartmentalize that and balance it, Margaret? Because I think that's huge for other founders to hear.

Speaker C: Yeah, I think at the core, and I'm going to go like existential, I'm going to stay away from tech for a minute because I think there's always a technical solution. But the harder part is starting to figure out like to your point, what is it that you're actually, what do you need? What are you asking for? How do you bridge that gap in insurance? I think especially right now with our climate, what distributors in particular and carriers are really asking for is, you know, transparency. In a lot of ways there is an uh, inherent distrust amongst all the parties in an insurance value chain. If we're going to be candid about what's actually going on. Customers, insured, people, they're frustrated. Rising rates, you know, increasing economic and weather related, you know, anomalies like uh, like everybody's under stress. And I think at the end of the chain people feel the, as though insurance is not transparent. They don't understand what they're buying, they don't think the coverage is what they thought when they purchased it. Uh, every time you have to get in a claim, you realize all the red tape that you didn't understand cause you're not a lawyer as an actual insured and you and I have both. I've recently had to go through the claims process. You had a terrible claims process. And we're in insurance and it's still difficult for us to navigate. So I can't imagine somebody who genuinely isn't entrenched trying to figure out why these things are happening. So get the very base level, everybody. Insurance wants to ensure a risk correctly, make it profitable and have transparency so that their customers don't have this intrinsic feeling like they're being ripped off. Like at the core, that's what we all want. And uh, people are asking for that. So they want to be able to see their customer, they want to be able to have the customer data and then they want to be able to utilize that. They want to know what's right and it's correct and it's traceable for underwriting reasons and, and then they want to be able to use that Information in their daily operational activities, whether they're binding policies or servicing post bind.

Speaker A: Right.

Speaker C: Like all of these things kind of drill back into did you sell the right policy to the right person at the right time with the right coverage? And that is uh, the core of what we're all looking for. Not the tech solution, but the core of what we're looking to build.

Speaker B: I think what's interesting about that, and I'm just going to kind of weave in here is because I think to your point, everyone's after that. Like.

Speaker C: Right.

Speaker B: Like I think in general. Yep. You know, we're an industry built on an intangible promise, uh, trust and creation of trust and transparency. I don't think there's many people would say they're not trying to do that. But then it's kind of interesting is because like, what line did you have to cross that others weren't tackling to get that? Because I think I call it like the treadmill or you could call it like, um, like a revolving door. People all start with the kind of core business objective aligned in principle. But I think it's like we go nowhere. You know, I heard you say APIs to nowhere. The revolving door, the treadmill, whatever. How did you decide and what line? So how did you decide to cross the line and what line did you decide to cross to kind of, if you will, stop the performative nature of what people were doing to achieve that transparency and trust?

Speaker C: Yeah, for us it was the human in the center design, I think that that led everything. So if we were really going to be able to deliver auditable, to your point, immutable, like uh, verifiable information to all the parties on the chains. I think all of us realize that there's a data exchange problem m at the heart of this. The next layer up is great. Now we have the systems and the data. How the hell do we give it, how do we exchange it between all of these disparate parties and all these different interested parties that take place in a policy. So I think for us it was, well, we have to agree to tackle the first step, which is can you even get that data in a place where it matters? And to us it was, I need a singular view of a person, whether they're participating in a, uh, corporation, whether they're standalone, whether they're in a household, a marriage, you know, however they're participating in our relationship, how do we understand that person and then how do we get all of that data surrounding that person to give us a really cohesive view of what not just their value proposition is, but what is their risk? Like, what, what are they, what are they actually doing and how do we, like, actually cover the risk at a much broader level? So for us to do that, like, and if we get into the tech side, uh, like rebuilding a boring basic relational database like just like every other company has ever done, it just, it wasn't going to garner the results that we were actually looking for. And I think for us it was going Web3 and looking at components like Blockchain to try to figure out is there something there that we have not explored before and we know other companies aren't exploring, at least not yet, and how do we actually make that a reality? And so that was a very big jump.

Speaker B: I think that that's so pivotal because, and clearly I've had the, you know, the specific perspective on this, of being very much in the trenches with you on this. So I have a depth that maybe our listeners don't have. But it was so easy to just performatively do what others did and try to do it incrementally better. And I would all say incrementally better, Margo. Like, you could have done what other people did before and you could have put a, uh, prettier UI on it. You could have made it feel Prettier and more 2025ish. Right?

Speaker C: Like, you mean this, this decade?

Speaker B: Yes, like this, this century. No, I'm joking. But, but, and you know what's really interesting? And I call that like the marketing or the performative nature of it, by the way. There's nothing wrong with that. Like, if you want, if people want to go do that, like, that's cool, Cool. But that is the treadmill and that's the revolving door.

Speaker C: Well, it hits a, it hits a wall. Lisa. We've seen it, uh, over and over and over again.

Speaker B: How did you know? Because what I think is so hard is most people, the inertia of VCs. And I'm not criticizing any VCs, right? Get to market, get revenue, you know, put training wheels on it, get it out there, get it used. Uh, that is the, like the undertow of the current. You're swimming in with money and funding. Right? And then there's the, like, I'll call it like the oh, shit moment. I could just do what everybody else is doing and make it a little bit better and probably sell it, by the way. And then there's like the whole reckoning which says, yes, but I'm not going to actually achieve what's necessary. And I think holding all of that in, in essence incongruency with the tension is something that most founders can't do. Margot M. Like, they, they have to like really difficult. How do you do that? A lot of whatever you choose to decompress, insert adjectives.

Speaker C: Well, it's, it's a del. It's a balancing act. So one, I think we had the luxury of building a lot when we weren't VC backed. So I made a lot of mistakes with my own money, frankly, right early on in my own agencies and our own brokerages. Like we, like, we hit that wall, um, with this human in the center and verifiable data. And how are we gonna actually show all of these relationships and how are we gonna get the data in a place that's usable? We did that early on with what the platforms that everyone else is building on and I won't say them salesforce, but like we, everybody's been down this road and we, I hit that wall

Speaker B: and let's throw some magentic AI in there and make it even faster to hit the wall. Let the crash dummies hit the wall in the car.

Speaker C: Whatever buzzword you want to put in there. The underlying operating system could not handle, like it just, it could not handle the level, first of all of, of data that insurance companies actually have. Like, if they were to represent their data digitally, which most of them are not, full stop. Like I don't know what the percentage is, but I can guarantee you it's pretty low, especially on the distribution side as to the totality of the data they have versus what they have actually digitized and is usable is like obscenely low. So in that scenario it's like if, if you're really going to put all the data in and you're going to digitize and you're going to utilize AI and you're going to share data across multiple entities. Like you have to have firepower underneath that that will support that or you will break it. And I think early on in our agencies, even the size we were, which we are not a, uh, large agency, you know, we're not, we're not the marshes and AONs and like the huge agencies of the world. And I hit those compute processing ceilings. I can't even imagine what, you know, a large carrier or broker is going to hit in that scenario. It wasn't cost effective and it wasn't going to work. So for us it was like, okay, we built and failed a lot in private. And so I was very candid with our investors when we took VC money that this was not. We, we had no plans to flip this business in three years. This was not a widget that we were going to build up with no backend and then exit to one of our competitors. That was not the plan. It will never be the plan. And so having money invested that understands that, I think was really, really critical for, for us.

Speaker B: So talk to me about the next layer before we get into. Because I want to go into kind of some of the provability points of Iris. But before that there's also. So, so it's kind of inverting the data model customers at the center. And I would assume that most naive listeners may think, oh, that's what CRM does. You know, cut. You know what I mean? We all have to go there. But okay, we've heard, we've heard those words before. So I want to go to another layer for, for you, which I know IRIS does. It wasn't just customer at the center, it's customer at the center. And with customer at the center, the

Speaker C: ability to do all of the operational processing. Yeah.

Speaker B: And I know that that's incredibly like an afterthought for most people that are thinking like, whatever, I've got HubSpot or I've got this CRM tool, or I've got that CRM tool and I've got like a database. I'm like, okay, like a database doesn't process. Like so, so like that operational processing. And I mean, you've heard me talk about this a bazillion times. I was like, oh, well, my reinsurance background, familial. When I went into the distribution background, because clearly we do policy movement and we do these things which are deeply, operationally intense. How did you combine that and how did you think about that as you were building this with, you know, your API?

Speaker C: Like.

Speaker B: Because then I want to get into your. Like, yeah. Why do APIs go nowhere? Like, like, how did you think about that? And what's, um, so important? And I think it gets back to the data exchange. But I'll turn it over to you to kind of bring the readers in on that, listeners in on that. It's also an actual operating system. It's not just a storage database. Right.

Speaker C: So I think people. Because let's just. I'm going to stick in, in distribution. Cause this is where I live. So we have something called an agency management system, an AMS or bms. Depends on where you are in the world or where you're listening. And so those words are really deceiving because it sounds like an opera, like it's, it's named as though it's an operational system. But what it actually is is a filing cabinet. It is a data repository. It is a data repository that was always built to be a data repository. And bemoaning of the UI like that's what it was for. It was meant to. I have this file and I need to digitally type into a screen the information from this file so that I can store it here digitally in a filing cabinet. That is if you take the words AMS or BMS out of the equation, that is what those systems do. And that's great because you know, in the 90s, 80s and 90s, like we need, that's what we needed, that's what we had. But unfortunately a dummy black box data storage software is not built and will never be built to run policy movements to, to do long, like long tail processing. That's actually, if you leave insurance, that's another separate system. That's your Oracles and your Sapiens. Like those are big processing platforms. And so if you were outside of insurance you would have. And the crazy part is I think people look at distribution as a sales organization and they're. And we are sales organizations but actually we're so much more highly complex because we don't own the asset or the premium that is the policy.

Speaker B: Right.

Speaker C: Like it's not our premium, um, to collect, it's not ours to run risk against. Right. But we still need to m be completely tied to every one of those movements. As a broker, like if you're really going to do a good job as a broker, you have to understand when a policy cancels, why it's canceling, is it for non payment, is there material misrepresentation? Like those are policy movements endorsing, canceling. And the systems that are databases do not have the ability to actually run the process now they can record the process once it's happened. So if you're an AMS user or BMS user, you're very familiar with something happens somewhere else. And then I go into my repository and I tell the repository that something has just happened that is not operational processing, that is record keeping. There is a big difference between the two and I don't think there's enough education for. You're trying for a lot of these bigger brokers too. They're trying to scale to operational processing like a carrier or reinsurer would need. But they're still using data repositories and they they don't understand why they can't get there. Well, you're not using the right software, like full stop. You're not.

Speaker B: Well, let's take that then, like, because I think this is such a natural segue. Iris embeds trust because you started with like, what's the holy grail that you were solving for like transparency and trust, and you chose to embed it at the actual data level. So you don't just reposit data, you're actually natively embedding trust. You know, we could talk about whatever I call all the, all the ways you achieve that in the like, technology world. Let's put that over in stage. Right. But at the end of the day, the way you achieve that is over here. But what you're actually doing is you're embedding trust and logic at the, like in, like, at the substrate level. At every single data object is, you know, individually, like provable.

Speaker C: So verifiable, exchangeable. Right. That is a massive shift in infrastructure software from what we have today.

Speaker B: Exactly. So when you did that, I would say, why did you do that? What are you opening up with that to your point about like, hey, I've already broken this, I've already seen the crash or the tidal wave that maybe other people can't see. And then what does that open up for Iris and Iris users? And then what breaks? Like, what's the tsunami coming for people that are trying to still operate at this data base storage level that you articulated, you know, filing cabinet, I love that word, the filing cabinet level.

Speaker C: I mean that's what it is. Everybody's familiar. Yeah. A lot of it is you are a company that relies on, you know, your ability to prove and to be trustworthy. And if you, your entire systems are not able to be audited, like, think about the absurdity of that. Like my whole company, my billion dollar company depends on my ability to collect the right data and then transmit that data effectively to all the other parties so that we can then find a policy and, and I can have my, my cover, my, you know. Right. And like you don't, your system just physically won't do it. So how can you possibly be human centric, transparent? How can you be good at underwriting and risk management and you don't even, like, you have no tools and no ability to do that. To me, that was just such a, like the chasm was so wide there. So for us it was a couple things. One, I mean, anybody who's ever owned an insurance company who has been on the gone to the altar at least and had to have an E and O claim. And most of us have had large ones and sometimes very, you know, frequently, depending on the type of insurance you write and you get audited or you get audited or you get subpoenaed. Man, like, think about that process and what we're going through currently and how that looks and our inability to, as brokers and distribution partners leverage any kind of buying power against a Swiss RE or a Munich RE when we're talking about REO exposure. And like we have no leg to stand on. And you being from the reinsurance side, I think the sentiment is. And I don't know if it's right or not, but it's certainly justifiable. Brokers don't have the sophistication nor the reliability of data. So what happens by the time it comes to you on the reinsurance side, Lisa? You guys are verifying, you're reverifying. Like there's billions of dollars in that process.

Speaker B: I was at a lunch one day recently and I said to this very large broker sitting across tape from me, you realize that like 80 to 90% of your commercial data you submit is ignored and we go get our own data. And he literally, Margaret looked at me, he's like, you're lying. And I'm like, there were two other reinsurers that worked at my peer competitors. Like at the table with me, we were sitting in a triangle kind of. And they're like, no, she's like, not lying. He was like, it was like, why

Speaker C: are we collecting it?

Speaker B: There was like this awakening Margot that he was like, what? Like he couldn't believe it, could not believe it. And I'm like, yeah, we go get our own. We ignore, replace, extend, backfill, verify, you

Speaker C: know, all the things.

Speaker B: Because we have our. We've just kind of gotten into like a self serve data mode. But to your point, it's about trust and verifiability. And we're like, well, we'll just, it's

Speaker C: kind of like we don't trust you. I mean the bottom line is we don't trust you. And you know, it's like, well, yeah, I mean that, that's the reality of the situation we're in. But think about as if I'm a large broker and I'm looking for my competitive edge and, and my value proposition and I want to be a partner. So let's say that I want to develop new programs. I've got a lot. I want to take the data that I have and I want to be just as reliable as my carrier, my reinsurer partners. And in order, maybe I want to take those programs directly to my reinsurer. Maybe I want to start my own reinsurance. Like these are real things, your own

Speaker B: risk appetite network or what, whatever. I mean there's so many risk solutions. Now. I agree with you.

Speaker C: I think the problem is, and I think a lot of big brokerages are finding this out and as it's moving down to mid market brokerages, you are woefully, woefully unprepared to step into that realm. Even though the return on investment is huge, the upside for a broker to be able to do these things with their data is massive. The downside is if like we makes our, we make ourselves obsolete. So Lisa, if you and every carrier in the world does have to re verify everything we're sending you, then at what point do we like work ourselves out of a job because we're unwilling to develop technology? Like at what point do you guys, and we've seen this, we've seen this over the last five years, say we'll find a different distribution channel, one that is reliable, that we don't need to pay 30% to because we're not only repaying you, but now we're re, now we gotta go redo everything that we're paying you to do, to have. And it's a, it's a contentious back and forth between distribution carrier and reinsurer in the broker space because it's been disjointed because there hasn't been somebody to come in and say we're going to lead the pack here and we're going to do these hard things. We're not here to just turn and burn and make a dollar off of brokers. Right. No one's led that pack. So there's no buying power in the broker space and they haven't been able to unify behind any one piece of technology. And unfortunately the ones that are in the market do not have a financial interest. Well, because why, why would anybody push the technology forward when we're fine doing what we've been doing for 30 years? It's a really, it's a quagmire and it was one that I think nobody that doesn't intimately understand insurance could possibly walk in and understand and then build against. And even those in insurance don't necessarily understand the level of technology available and, or needed to actually solve that problem. So it's, it's a hard space to be in.

Speaker B: You know, for me like seeing, seeing so much of what you've done. The next level to me is like, it's that uh, how did you create the data and how do you create the transparency and the trust? It's also how do you create and I'll use the word dynamic, but how do you not do period and close? Like how do you get what we would call more event driven even if you're not 100% event native at the outset? How. Because you know, whatever, there's still things like downloads, there's still things like you've got to go to a portal.

Speaker C: Yeah, we're batching some things.

Speaker B: I mean let's call it hybrid mode right now because you know, you can't build for something that the other side isn't meeting you at either.

Speaker C: Correct, correct.

Speaker B: But like, why is it such a difference to not have something like a period and batch? Like, like financial engine? Like, like when you designed Iris, it's really like producing. Like I can see in real time my margin, I can see my profitability, my producers can see that. How is that different from like this concept of like type in all my transactions? And I'm going to like, like why is that so important to run the business?

Speaker C: Because how like our uh, world is not, we're not making decisions by quarter anymore like, or by year. Right. Like it's funny actually because I'm here right now, I'm recording this with you and I'm in San Francisco, I'm here in Silicon Valley and I am meeting with investors. We recently had an investor from South Korea and I'm meeting with a bunch of Asian investors that are in insurance. So you know Ms. Nad and Sampo, there's a bunch of really great Asian insurers and I've been having discussions that I have found just so starkly different from how we think of things. Other countries, other investor types, they talk about investment in terms of generations. They're not thinking what is this quarter's numbers? I mean everybody is. Right. But the rest of the world is thinking in generational, like what are we investing in now that's going to do return in 10 years. To me that's something that I see in the, on the broker side in the mid market, these great legacy mid market providers that, you know that have built something that are continuing to do M and A. Right. These really. And they have this great opportunity now to kind of keep going. Those people are thinking in generations, they're thinking about perpetuation plans and how they're moving things from their, you know, to their children and all of that. And so those thought process are a lot longer. But if you move into some of the larger brokers or the really small ones, it's always this short term planning of we got to hit this quarter's numbers and this year and like, well, we think about next year once a year. It's a difference in minds and mindset shift. But the ones that are thinking in the longer term, generational planning, they're winning right now. They're the ones that are winning all the M and A deals. Right. Because nobody wants to be a part of a turn and burn. Everybody wants to be a part of something. And when you think about what it takes to plan like that, it uh, like you need to understand how to make business decisions in real time. You need to understand when a market is shifting. You need to understand when an insurer is moving or a reinsurer is moving. And you cannot do that if you're doing a one year annual planning on what next year will be. You need more information. You need the ability to pivot, you need the ability to quickly disseminate from the top down an initiative. So maybe this year we thought we were going to go after ag, right? Because this is like a really great market and blah, blah, blah. And then boom, all of a sudden immigration happens and now, oh my God, AG is actually falling apart. Now. How do we take this initiative that we planned for and how do we move it quickly? You can't do any of that when you don't even get to the numbers until 90 days after something happens. Like how are you actually going to be nimble enough to respond to a world that is in real time?

Speaker B: Yeah. And I think part of the fascinating thing for me, when I started working in distribution with you and started understanding, I think everyone thinks they have that capability and the way in which they define it is, well, we'll send out an email to all of our producers and I'm like, what? They don't have the ability to. It's like a leak, you know, like that automatic faucet shut off that shuts the water off and it prevents the leak from getting like immediately. I think of that with capacity and risk placement. Right. You've got to be able to switch your flow immediately. And sadly humans aren't conditioned to just read an email and always do what the email says. If they can go in and do whatever they used to do, they're going to keep doing it.

Speaker C: Yeah, it's like pattern. They're like trained. They got muscle memory, right. You're breaking that muscle memory. And it's really important that like I cannot stress, like if you talk about, like we've talked about what's coming after this, like we lay this foundation layer and you get this real time and you get this verifiable data, like what does that mean to you as a business? Yes, it's nice. You're, you're going to have pivot, you're going to have the initiation, the um, you know, all of that's going to happen. But what does it mean long term? Well, if I, let's talk about AI for a minute because that's what everybody's talking about. Even though I hate talking about it, it's just, it's over. It's, there's a lot. But if I am a big broker and I now have, my board comes down and says, hey, everyone's using AI, like right, this is the, this is the buzzword in all the, the boardrooms. We want to deploy AI and you don't have verifiable data to deploy AI on top of or you don't have an operational process to, to insert AI into. How are you putting, yeah, like where does AI live on your data? Like in your filing cabinet. Like, what is the best AI can do with a filing cabinet filled with broken, wrong, dirty files? I mean, what is the best hope you can get unlipped? I think you're seeing this right now. I mean you and I have been in so many of these big broker rooms and they're, and they're like, we got a pilot, we got a pilot, but nothing rolls out and nothing's returning the investment. So these executives and boards are like, well AI doesn't work. It's like you're trying to deploy something on top of something that is never going to, it's never been built for, it can't support and it also can't feed. And we're looking into the future now with E and O where you're going to train. This is happening right now. People are going to start training these LLMs on this really bad broken, non processable data and they're gonna start getting into big trouble, right? You're gonna, your E and O is gonna go like, so it's gonna be who can figure this out versus who can't. And it's not an easy problem to solve. And if I had a dollar for every time someone was like, well I just gonna turn GBT and I put my M coordinate into it and I'm like, okay, but like cool, and that's great for writing an email, but like I'm talking about operational processing data at speed and at volume and how is your LLM gonna respond and how are you training it? There's no concept of what is actually needed to make that successful. Well, there's not.

Speaker B: And just to amplify, ah, that anyone, and I will say anyone, that has ever been involved in any level of any enterprise grade technology, like you've worked at a carrier, or you've worked at a corporation, or you've worked at a reinsurer or whatever. We never used a CRM system, Margot, without connecting it to some sort of master data management or data harmonization layers like we had data governance. So if you've been removed from that and you've got your data in this I'll uh, call it CRM only, database only universe and you haven't connected that. Throwing AI on top of that is by definition asking it to hallucinate. You're like inviting it to a drug.

Speaker C: Yeah, you're, you're training it to hallucinate at that point.

Speaker B: Yeah, like that's what you're training it to do. I mean even the organizations that have had decades of governance, they're still not easy to implement operational AI to your point, but they're getting there. And so I think that this disconnect of um, I can just throw a chat GPT on top of my database and everything's fine. To your point, that's naive. But what I find even more interesting is that we're not even proving that. And then we're like, hey, let's add some agents in there and go full bore on a jet. Oh my God.

Speaker C: But yeah, let's, let's go full box.

Speaker B: What do you think about that?

Speaker C: Yeah, let's do that real fast and at scale if we can even get it up and running. Nobody has any business talking about agency AI if they are still operating on a database with no master data, no intelligence layer.

Speaker B: My favorite part about that is you can't even do agent to agent communication vis a vis an API. Okay, so let's go back to the world of API.

Speaker C: You got to put all kinds of layers in between that.

Speaker B: Most of these things can't even do API to API communication that's not event based. Now we're going to go into, we're doing agentic AI and if anybody listens to this, your people are telling you they're doing agentic AI and all you have is at uh, best API calls hashtag no, because an agent has to talk in an agent to agent. So there's the Google agent to agent, there's the managed control plane. And we are only seeing very sophisticated companies start to roadmap out that. But Margot, the amount of distribution people that I hear saying they're doing agency and I'm like, they're not.

Speaker C: I mean, look, and there's nothing wrong, like if you can rig up something that, that works. Like I'm, um, like more, I've done it 100 times, like more power to you. But you need to understand the repercussions of what you're potentially building. And again, I think about, there's a couple instances that I've been a part of where this uh, isn't even agenic, just API. When an API call, when you're pushing 8,000 quotes through an API and it's coding NIC codes, NAIC codes incorrectly, and you just wrote 4,000 policies that now all have material misrepresentation. Right?

Speaker B: Yeah.

Speaker C: And a carrier accepts.

Speaker B: That was literally like a bot, by the way. Yeah, yeah, that's a bot.

Speaker C: That's not a genic. Right. So like you're talking about even taking that and then a carrier doesn't even realize it until 120 days when they're doing their post audit, which is also happening by humans going through, looking at things. And uh, now all of a sudden we've got millions and hundreds of millions potentially and just complete, utter chaos. I think about that and we don't even have a genic yet. Like, we're not even talking about a genic. And I think about the repercussions of that. And so I think, again, I think it's really naive in two ways. One, to say that we're at a place that that's even deployable, because I don't think that we are. And two, that even if you could rig something up that you understand what the and O repercussions are of that build, they're massive.

Speaker B: I agree with you. So let's get into like the inside the build. We'll keep it short, but you know, I think a lot of times we talk to founders and it's like, you know, it was hard, triumph, success, but it comes with a huge cost. So one of my favorite things to reflect on is like, what almost broke you and what did you learn from being on the edge of failure? I mean, you've never had a safety net in anything you've ever done. Why haven't you walked away? What is your pulse point that keeps you on the edge?

Speaker C: I Don't see anybody else doing this. I don't see it in my peer groups. I don't see it when I'm on stage. I don't see it when I'm behind the boardrooms with these executives. Like, I see the problem is large in scale, it is very difficult to solve. And I see a lot of people come in and enter and then I see them realize that it's going to take literally every ounce of self preservation just going out the window to even attempt it. And I see that. I see them, I see their face turn and I see them give up. And I cannot imagine knowing what I know, like and being so passionate about. I know how to fix this problem. Like we have the knowledge, we have the path forward, we have the technology. Like it exists in the world. To have all of those things and then to not be able to get this industry to where it will, it's going to go there at some point. Like there is no other path. It has to or it will fail and, and to walk away. I, like, I can't even imagine a scenario where unless I was forced to, I would never, I could never.

Speaker B: Yeah, I think that's such a, such a strong part of this journey that we often don't talk about because I think, I think a lot of people, and I'm not disrespecting any founders, but they feel a calling for an exit or until it's hard.

Speaker C: Uh, Lisa, listen, I'm a VC startup. Like if you don't think that that has been uh, like kicked around and talked about and all, like you have to. Because one of the things that I think I bucked the system on and I get so much shit for it online. You'll see it everywhere. It's Iris is vaporware. It's never going to launch. Is the idea that I'm not building towards ARR in a sense so that I can flip this company. Right? And there's this. I call it like the demo. I call it like every founder gets to a point where they build towards a demo.

Speaker B: Yeah.

Speaker C: So they'll go. Because everybody's like, oh, I don't understand. I want to see the system. Let me get in and play with you. Like. Well, it's an operational system so like, good luck. I mean it's triggering operations across, you know, 42 different platforms. But yeah, let's set up a demo. Org. But you have to take.

Speaker B: Here's a sandbox for that.

Speaker C: By the way, everything in the sandbox is completely fake. Right? So like it's a, it's like a, it's like a um, like a safety net or like a baby blanket. You're like I got my demo work so I feel really good about this. It's like everything in there is just completely fake and it's all being triggered and hard coded and it means nothing.

Speaker B: It's like a simulator. Yeah, yeah.

Speaker C: I mean but which is okay. Like if, if and you know we'll, we're on the path of building the simulator so. Because it is difficult for people to understand. A lot of people are visual so they need to see this, what this process means to them and that there's no shade there. But that was not what I was going to spend our limited resources and money on building a baby blanket for people that it doesn't prove anything to. I think for us that was, we take a lot of heat for that but I stand by my decision. And we have customers that are on the platform that have access to the system and they're clickety clacking in there right now. But it doesn't mean that you know, we spend a million or $2 million in an irresponsible way to make people who probably weren't going to buy the software in the first place feel better about the software.

Speaker B: I mean I think it's interesting, it's interesting juxtaposition because there are very, very, very large enterprise platforms that do not have an environment that you get access to or you get to go play with.

Speaker C: I mean once you're post, you know, if you're in an implementation you get a UAT or you get sandbox, that's normal.

Speaker B: You don't get anything till you pay. Like in most. Let's take Oracle, Oracle, ERP or Workday or like let's just use something like that so we don't like throw anybody else under the bus. That's a, like you sign a multi year contract, you get an environment right like this. But it's really hard to be, I'd say a uh, you know pre series a founded startup and say that. But again that comes back to you're building an operational system, not a widget. And I think widgets can have sandboxes. I think operational systems don't. So I think that's really interesting that you've had to hold that tension against.

Speaker C: Oh it's really difficult. I mean it's uh, a, it's very difficult and I think look rightfully so because especially on the broker side so many people have come to this market, built something that is really not stable or really doesn't have anything behind it, has sold it for a couple years and then flipped it, only to leave a whole bunch of people that have invested in it and spent time in it, you know, with nothing, and they're right back to where they started. So. And I was, uh, a victim of that as well. So, like, I fully understand where it's coming from, but I think the, the understanding from our market needs to be. When you're building something that has never existed before, never existed before, and you're going to market with it, it's not going to be for everyone, but it'll be five, six years before IRIS is big enough where we can deploy it.

Speaker B: And crossing the chasm, right?

Speaker C: Crossing the chasm, exactly. And we cannot do that out of order, as you say, we cannot. So if you're, uh, you know, if that's the. If you're looking to be someone who is not first in, Then don't look for first in software. But then you have to understand that you are not going to get first in results. Like, that's the whole part of being the leader of the cutting edge and being that kind of personality is it does not come without risk inherently. Yes, there's risk there, but you cannot expect extraordinary things if you are not willing to be extraordinary. So if you are the kind of company that wants to be. To have a hundred companies go before you, that's fine. There's lots of successful companies that do that. But you are not going to get the results that the first hundred companies got. Yeah, as buyers, I, uh. You got to wrap your brain around what that means for you.

Speaker B: I love that. Well, let's go into. I mean, you and I could.

Speaker C: I think we could do a whole

Speaker B: nother podcast just on. Cannot cross the chasm out of order. Maybe we should do that one for like, one of the holiday weeks to get a lot of attention. I love talking through these things with you in a format where other people can hear them and benefit from your journey. Let's go into my final call to action question for you, which, you know, I ask every guest, but it would be like, not an end of an episode if we didn't have this question. So, Margot, with everything you've said, with the journey you've been on, with the moments of tension that you have to hold with the market the way it is today, with the actual problem you're trying to solve, which is this kind of reckoning with capital trust, human in the center, actually honoring and delivering to our customers what is the thing that you wish everyone listening would start doing, stop doing and continue to do.

Speaker C: Okay, let's see. This one always stumps me. I think what I would advocate everyone to start doing is to, to start being honest about where they fall in the technology journey and what their, you know, threshold and capacity is for risk and then understanding that as a company and then applying your technology strategy to that. Meaning if you want to go fast and break things, then you need to be looking for companies that want to go fast and break things. If you are not that company, then you, you gotta stop being upset when you try to buy a uh, product from a company. That, that is, that's my first thing. So like start really being honest and self reflective about where you are in your tech journey and being able to own that with precision. I would stop using, oh God, this drives me, uh, nuts. I would stop saying that you're using a genic AI because you have a handful of automations that are running sequenced. That is not a genic AI, please. And I'm not saying that to be rude. I'm saying that we need to understand the difference and the severity and the difference of cost and E and O exposure to what we're actually talking about when we say agentic. I think it's amazing that people are even talking about it. We weren't even talking about generative AI two years ago. Now we're all the way into agency. I think it's wonderful. But let's stop misusing the term because it's not helping us move forward in the actual deployment of that tool and then continue. I'm going to continue to advocate that the insurance distribution channel is worth the time, effort, energy. It is a valuable part of our value chain. I think it might end up being the most valuable part of the value chain at the end of the day since we hold the relationship with the consumer and that is our strongest asset. And I would continue advocating that we deserve the kind of technology that other players have and that our size and our current education level should not preclude us from being involved in the discussion moving forward.

Speaker B: I love that. Well, thank you first of all so much for being a guest today. Thank you for your grace under fire in the hot seat. As always, for anybody that doesn't know or follow Margot on LinkedIn, you can follow Margo Giles on LinkedIn. You can also follow Iris Insurtech on LinkedIn. She and the team do just an amazing job of, I uh, would say like no nonsense, truth telling, keeping it real. Education constant education like of, of the industry and, and lots of, I would say uh, webinars, awards participation. I mean you can just feel the energy whenever I look at all of the posts and activity going on on

Speaker A: um, your social channels.

Speaker B: So thank you so much for being a guest today. Margo thank you Lisa. And to all of our listeners, as always, stay curious, stay informed and stay plugged in. Thank you so much for being a guest.

Speaker A: Today's episode of Insurance Unplugged, the AI and distribution series is proudly sponsored by Iris InsurTech, your gateway to the future of insurance distribution. IRIS harnesses the power of generative AI to transform data processing and decision making across the distribution landscape. The IRIS platform integrates AI driven decision engines, dynamic form generation and configurable workflows all underpinned by continuous data quality management discovery. How IRIS is powering smarter operations and more efficient distribution with cutting edge AI setting a new standard of excellence across the entire industry.