Theo Butt & Andy Roberts: Kinetic Insurance Services and Guillaume Bonnissent: Quotech: Building an AI-native MGA from scratch: inside Kinetic Insurance Services (419)

0:08Welcome to this week's Instec podcast. There's a whole gang of us this week. I'm joined by Theo Butt, who's the CEO of Kinetic Insurance Services, Andy Roberts, who's the CTO of the same company, and Guillaume Bonnissent, who's the CEO of Quotec. We've assembled this crew because we want to talk about Kinetic Insurance Services, launched in June by Convex. So this is the A-Team. Thank you all very much for joining me. Thank you for having us. I'm going to start with you, obviously, as the CEO. Convex launched an MGA business that sits alongside your underwriting business. It's intrigued people a great deal.

0:51We are desperate to find out more about it. tell us what is Kinetic Insurance Services? Thanks very much, Robin. So yes, Kinetic is something we've been planning for a while now, and it's a core component of the next phase of Convex's growth and the iteration of the business as we think about the next five years. And really what we wanted to do was to find a way to increase our access to addressable market, stretch our risk appetite, and do that utilizing the strength of Convex, the carrier, and also by accessing third-party capacity in order to give us relevance. So that's where this idea came from.

1:25And we spent some time thinking about how the best way was to do that. And this hybrid carrier MGA model was what we landed on as our solution. So we've spent the last year really putting the team together for the operational and technology components of that. And really, it's a fantastic opportunity for us to build again from the ground up, utilizing latest technology, agentic AI capabilities and genuinely building a very forward-thinking and tech-focused platform. So yes, it's a platform for us to increase our access to lines of business that are adjacent and complementary to what we do in Convex, the carrier, and we're very excited to be here to talk about it today.

2:04Then what kind of underwriter, what kind of entrepreneur are you hoping to attract into Kinetic? So the advantage of building an MGA with support of a carrier is multiple. So firstly, we have access to capacity. So that's a big advantage when you're setting up your own MGA. We hope to be able to attract entrepreneurial underwriters who want to step into an MGA model, but will also understand that we can provide them with firstly that access to capital much more easily, so they won't have to spend as much time going out there looking for that, but also the infrastructure that we are building around them.

2:43So a cell for them to operate within and be entrepreneurial, but we will put the tech platform in, we'll put the operational platform, the Quotech system around them, as well as having the resources and capabilities of the carrier to call on from a claims perspective, compliance, and all of the other component parts. So we want to attract people who are excited about building a new business for themselves alongside us as part of the Kinetic brands as we grow that collectively into the future. Andy, that brings me neatly to you because Theo's pushed hard there on the technology benefits that Kinetic will provide to entrepreneurs.

3:19You've got the job of building this out. What is it that you are going to provide your entrepreneurial underwriters as part of the kind of tech infrastructure they'll get when they come on board? Yeah, thanks, Robert. So, Obviously, we have to stand up the core operation, and that's what we've been doing over the last nine months. And that is with a number of core systems of which Quotech is one. But we're also doing some really exciting build elements around it as well. So our vision is to eventually move to a world where almost all of our non-revenue generating work is automated, and also to bring the insights to underwriters so that they can generate more premium.

3:53And how we're doing that, we partnered really closely with AWS and their SI Perfectus to help us to build an agentic native factory. So this is effectively our core processes, all the way from the first email coming through to generating our quotes with the underwriter very much in control all the way through this. And this is an opportunity we saw as agentic technology was now available to us as we were starting up. It's also a greenfield world that we can play with. So we've partnered, as I say, with the hyperscalers to use the latest tech to really build out an innovation platform that we can start deploying not just the core processes, but we'll also call superchargers.

4:29So all those annoying little repeatable tasks, we want the underwriters to come and say, this is what we want. And we'll have a team that can deliver that to them as quickly as we can. So the core components, if I've got that right, are kind of agenticized workflow. So then presumably that links into some core record policy admin system. That's the core of the platform, is it? Yeah, absolutely. And that is where Quotech comes in. So we made a real effort to find something that we could build with and around. So we looked for something that was built truly with MGAs in mind. We're never going to want to really build our own ledger.

5:08And through that, we wanted to buy something with a super strong ledger at the back end. And we wanted a policy admin system that covered everything from underwriting through to the billing collections piece and not just a shiny underwriting set of screens. So that's one of the main elements for why we chose Quotate. But behind that also, we needed a really open platform, something that's API enabled, something that's moving towards MCP enablement so that we could build around that core as quickly as we can. Guillaume, you must be thrilled. This sounds like a dream gig for a man who's been building platforms like this and preaching the case for new modern platforms.

5:42How are you getting on? No, it's great. We've been working with Convex since about the start of Quotech. It's always been a great partner for us. And when they started telling us about what they wanted to do with the MGA at the time, we didn't know of the name, so we called it a Convex MGA. We got really excited because we had perfect alignment between the vision we had of how platform should be used. And it reminded me 20 years ago when I was at Hiscox, we had a project called Underwriters to underwrite, which never really worked because the technology wasn't there. And we didn't have the core system to be able to do that.

6:19So underwriters spend a lot of time doing admin tasks. So this was a really exciting project, and we're very pleased to be involved in it and to have won the RFP. It's not a standard gig, is it? It's not like you would go to a standalone MGA and walk in and sell your platform. You can have multiple underwriters, multiple cells, multiple requirements. Have sort of any particular challenges from your point of view? Yeah, it's really interesting from that point of view, because at the moment, we don't actually know who the underwriters are going to be. We don't know what kind of complexity they're going to bring, etc.

6:54So we built our platform with that in mind. And since we started the project, also we changed it to be very flexible so that when Convex hires a new team, the technology doesn't become an issue. A lot of large MGAs employ a new team. They need to look for capacity. They then need to configure the platform to be able to work with that new team. And generally, that can take three to six months. We've really worked to reduce that dramatically to a matter of days or weeks. And that's been the really exciting part. Challenge if you want, but that's a challenge that the team really took on. So it's quite interesting because we don't know if the products are going to be some super complex reinsurance product or they're going to be something as simple as Goldfall in one.

7:40But the platform can do both and everything in between. Theo, I wouldn't be doing my job if I didn't challenge the model a little bit because there are aspects to it which have greatly intrigued everybody. So indulge me. The first challenge, is this a bit of a hybrid model? And what I mean by that is, won't really entrepreneurial MGA owners, underwriters, be brave enough to go out and set up their own business and seek investors and claim more of the value of what they create? Are they going to be compromised by the fact that you are part carrier, part MGA platform provider? Is it for those who are looking for a safe option, or are you hoping to attract the truly entrepreneurial?

8:31Yes, hybrid is one way to describe it. And of course, that is what we are. But that, as we described already, brings its advantages in terms of capacity and other ancillary services. But Kinetic is a distinct company as well. We've set it up as a separate company. It will have its own branding adjacent but separate to Convex. It will have its own compensation arrangements that are in line with MGA compensation arrangements. So we will be able to incentivize people appropriately to build their profitable portfolios with us. So I think in terms of that element of it, we should be able to compete with, if not better, any other MGA out there in the market.

9:08And then I think actually using Andy's phrase from earlier, it's another hypercharger. What we can do is put people into a framework that allows them to really focus on the value-added work that they do as an underwriter rather than having to worry about building systems, processes, back-end ledgers, all of the stuff that traditionally we're not very good at as underwriters ourselves. So I think it's the opposite of that. I think we will genuinely attract entrepreneurial people who then actually can get on with really building their business rather than doing all the mucky stuff around it and below it.

9:40That won't suit everybody. Some people will, as you say, want to plough their own furrow in its entirety and take their chances. We've seen lots of people do that successfully in the market. We've seen lots of people do that less successfully. And I think that's what we have offers a genuine differentiator to that model. And I think that's exciting. I think that's fair enough. And then are you in any way worried about potential clashes? You make a very good point earlier on about the ability to use this to extend your underwriting appetite and go into areas where you don't necessarily have existing expertise.

10:12But what happens if you attract people where you do have expertise? Are you worried about the two businesses, one Convex, one kinetic competing with each other? One of the sort of foundational early on principles is that we will be complementary to not competitive with what's been great. Convex is a very nimble company, despite its relative size in the market. We're still only 480 people. I know every single underwriter next door in the Convex building. we're working hand in glove with those teams to develop opportunities that they would support and where we find that there is a cross-section in the Venn diagram where there might be a little bit of crossover we'll work through those areas together and I've spent most of this morning doing that on a couple of potential lines of business that we're looking at so the nice thing is that everyone's supporting this initiative from within the carrier I stay as a group executive member so I'm very connected into what we're doing throughout the group of Convex myself and Doug Howard, the chief underwriting officer, sit down weekly to discuss potential opportunities, how they might fit, where we might best put them.

11:15So yes, there's a risk there, but it's one that we're well set up to mitigate with the way Convex is structured. And you mentioned capacity at the start. You will provide capacity. Do you guarantee to provide capacity? Presumably, if you don't like the look of it, you'll turn people away. But in reverse, are people obliged to take your capacity or can they come in and not take your capacity? How does that bit work? So I think from an alignment perspective with other third-party carriers, it's very important that Convex is enabled to offer cornerstone capacity, as we're calling it, on every cell.

11:49Otherwise, if we set up a cell and we go, we don't really want to support that one, why would anyone else either? So we will always offer cornerstone capacity. So the opportunities I bring in have to be done in a way that Convex is prepared to support them, either with lead capacity or support capacity depending on the line of business and our expertise to be able to lead that line of business. There may be the odd opportunity that come along where we don't provide capacity, where there's already a 100 % solution in place. If a cell comes along with their own capacity already set up, then we would obviously consider that subject to that capacity provider being happy to work with us on that basis.

12:26But foundational principle, again, is generally Convex will be a cornerstone provider for every cell that we're under right. Thank you. That's the end of my official challenges to the business. All I'm doing, I hope, asking the questions that others are most intrigued to get answers to. And it seems to me that you might have a sort of dream job, that you've got this Greenfield site, you can go and tap up AWS, you've got the chance to build something really exciting and really fresh within a large, well-funded enterprise. And now you've got the very best of AWS's agentic capabilities folded in.

13:07That raises the question with me, do you envisage situations where what you learn doing this swanky new platform might inform one day what Convex itself does? I think it's a great question. There's elements of what we're doing that Convex are already doing, and there's elements that we're doing that Convex aren't. So there's loads of learnings to be shared across it. I think the thing that we're doing, going truly build first in this, is looking at grown-up AI, which is the hard part, right? So how do we build something that is built for a regulated market that has clear explainability, auditability from day one, not something that you bolt on later after you've got it built out?

13:46So what we're trying to look at is building out, designing, and delivering out the patterns of how we structure and manage the context. next? How do we link in an email as soon as it comes through and route it to the right case? How do we automatically kick off those agents? And then really importantly, always see who's done what, when. So an agent has done this and why. Make sure that they have serious controls over what they can and cannot do. Keep a human in the loop the whole way through. So we are in no way looking to replace the judgment of an underwriter here. This is all about how we just bring them up the insights and to remove the kind of operations around it as much as possible.

14:23And then, of course, we've got all the fun later on about how do we then change the skill set. So the people right now in operations become more agent managers. We now need to build skills over there. Our underwriters will be way more AI native and tech savvy than historically. There'll be lots of learnings, I think, more on the operating modelers as well that we'll be able to share back over to Mothership. just be warned that we'll be tapping you up for next year's agentic ai in-stack event where you can come and share your learnings i should be greatly intrigued to find out what you're going on yeah i'm sorry i couldn't miss the opportunity automation has transformed individual parts of underwriting but connecting broker channels mgas insurers ai tools and core systems is a much harder challenge.

15:12On Wednesday, 7th October, join Instech and Duck Creek and SEND in London for Underwriting Rewired. Speakers from SEND, Duck Creek, Torch Underwriting and The Hertford and more will explore how to move beyond isolated pilots where implementation stalls and what must remain in human hands. Registration is free for insurers, brokers and MGAs. Visit instech.co to secure your place. Guillaume, how's the rest of your business going? Convex and Kinetic's not your only customer. How's business generally? Yeah, really good, actually. We have a lot more interest. I think what AI has done is really triggered a lot of questioning in different boardrooms about the systems they have.

15:53How do they get rid of the legacy system? How do they future-proof their architecture? And actually, interestingly, that leads to a lot of conversation where they realize that having a platform that is something as simple as API first, that's got MCP servers ready is much better to future-proof their architecture. And companies are now, again, realizing that the most important thing is to have the data. So how do they record all the data they require and how can they interrogate all of that data? And therefore, what is the architecture that matters to have that, be it a data lake and be it a write-on the writing system?

16:36That is where all the conversations we're having at the moment are going. You should expect in the next 12 months some really interesting announcement on our part, as well as obviously the progress we're making with Kinetic. You must have, from those conversations you have with the customers, some clear ideas about what you're going to have to invest in. Are there any sort of themes and topics to bring out of that? I think it's the integration with a large language model. How do you let your underwriters interact with large language model? And if you asked me that question a year ago, I can't remember last time we had a podcast, but if you asked me at the last postcard, well, I thought our users would log in onto our platform and see a prompt.

17:19Now is the progress in large language model and all the work that companies like Anthropic have done with B servers and making them open source. I think the first thing that's going to happen is every single user is going to log in and go on to their large language model user interface that's been given to them by their company. Why is it important that it's given to them by their company? It's because of all the compliance angle, making sure that they're not in breach of every single NDA they sign, making sure that the data is not being trained on, that it's properly segregated, et cetera, et cetera.

17:55Now, through the MCP, so what they can do is they can actually set up tasks. So for instance, let's say you sit at your desk at 8 a.m. every morning. Therefore, you want to see all the meetings that you're having on that day, all the tasks that you've got to do, every email that came in overnight. And then it could be things, if you're an underwriter, how many new inquiries come in and got automatically ingested into your platform? How many underwriting referrals have you got to do? How many queries from accounts have you got on things? And then you go onto the different platforms. I think things like going onto our, I reckon people are going to do that less and less.

18:34Does that mean that people are going to go less onto our platform? No, I don't think so. But it's just more the way they're going to access our platform is going to be through widgets on the large language model. Now, the way I describe it, I have to pinch myself to think, is it in three months? Is it in six months? No, it's actually doable today. And that's the thing that is extraordinary with technology is that all of that you can do today. Data ingestion is not a dream anymore. Having everything brought into your large language model isn't a dream. It's absolutely doable. So the question now is more an issue of adoption and also for IT departments and legal departments in different companies to work together to make sure that they don't say no one can use a large language model, but more look, OK, we have to use large language model.

19:27also which one do we need to use what package do we need to take and also control cost because the last thing you want to do is to have a build because every single user is vibe on the writing vibe coding and things like this so there needs to be all these controls in place but that is 100 the problem at the moment is adoption rather than where the technology is at that's really interesting when i said earlier that i was jealous of andy and all the excitement of his job all those controls and all those things to think about make me less excited when i think about it see over to you casting forward 2027 you've launched there's a sort of broad fascination and admiration i think for what you've done presumably now it's about signing up some mgas some underwriters anything else for 2027 what are your plans so yes really it's about finishing off now the final components the initial build and that's really just the day one model we're going to continue to build and iterate that and some of that will be part of what we do next year but from my perspective yes it's very much now about talent acquisition and attraction finding the right teams to join us to bring the complementary skill set and complementary underwriting appetites into convex.

20:47We've got three cells that we think we have up and running in time for the 1.1 launch with underwriters hired into those. First underwriter joined us in Bermuda at the beginning of August. So that's an emerging industry cell that we're setting up initially based out of Bermuda. So that should be up and underwriting in the next two or three months, dependent on regulatory approvals and what have you. And then we've got two further cells that we're working through. We should have three or four sales indeed up for 1-1. And then, yes, we have ambition to add at least two or three more sales during a concurrency of 2027, start leveraging all of the benefits of the technology that we're building by getting some people into underwrite it and start generating some revenue and some underwriting income to Convex and to our capital providers as well.

21:33So there's a lot to do. We've got to find the right talent. We've got to find capacity to support that talent. And we've got to plug them into our systems and processes. And then long term, is this the sort of business that Contex might spin off, or is that looking too far ahead and I'm just being presumptuous? So we have two MGAs, MGUs now. We built Convex US on the basis of an MGU platform as well, because that was the quickest, most efficient way to set up our boots on the ground operation in the US. We built that to be separatable. However, it's absolutely a core part of our distribution strategy to distribute Convex's capacity into indigenous US markets.

22:11With Kinetic, it's a slightly different strategy in that we are going to utilize more third-party capacity from the get-go to complement the Convex offering. But distribution strategy is different, and it's initially based here in London and Bermuda, as I just described. Eventually, we may well look to broaden that distribution base further afield as well. But we're building it as a separate entity, and it's its own legal entity owned by Convex. But again, it will form a core part of our next five-year distribution strategy. So when we start looking at premium, it'll be not just GWP, it'll be gross premium managed by Convex.

22:48And we don't have any intention to sell the platform. It's going to be part of our infrastructure going forwards. We're getting near the end. And my last question to you, because I can't work out whether it's a job that I would really like to have or whether a job that I would be terrified in. Are you enjoying yourself or do you find the responsibility of being given a budget and a sort of the access to everything shiny but deeply intimidating? What do you think? It's a constant state of both, Robin, very much somewhere between exciting and terrifying, for sure. It's the opportunity is massive.

23:25The technology is ever changing. The second you think you've just about understood it. Things move on at a speed of knots. I would say it's been a hell of a learning curve, but it's been an amazing experience and more on the exciting side than the terrifying side for now. And as I say, investing into those, what might be more boring stuff, but the reliability, the trust, the controls, the audit around AI will make this slightly less terrifying rather than the other way around. So I'd just like to also add that it is a big challenge we have in front of us, and that can be intimidating. But the team we've put together, both internally, but also working with the Quotech team and with Provectus, who are the delivery partner for Amazon, have made this a really exciting and enjoyable journey because everyone's really pulled together.

24:10People see the benefits of what we're looking to do and I think are quite excited about the prospect of building something that's really quite innovative and genuinely differentiated. So all of those partners have come together and have managed and sat over a number of large technology projects in my life. And I'm really appreciative and enjoying how this one's coming together. I agree. And I would also add, I've also seen some underwriters have early sight of where we are, and they're saying these are things they've not seen before. So there is a real excitement brewing around what it will mean for underwriters as well as they come and join that party.

24:43Of what it's worth, I love it in the sense that all your competitors are having to build, take this tool set and build it over some existing legacy capability, which gives them enormous handicaps in how they can deploy and the speed at which they can deploy. And people like me, who are able to just spout theory completely disconnected from the reality of doing business, have always been talking about that, why doesn't big carrier go and take a greenfield site and start from scratch and build this stuff out and get all the attraction of being able to do that and this seems to me a marvelous way to be able to get the best of both worlds and continue with what's a very well respected insurance business while you get to play with all the shiny new toys and the next two to three years is going to show the difference between people who've invested in this stuff and know how to make use of it and those who get left behind whether that be because of the state of their data or anything else i think just going back to an earlier point that one about adoption when we're not seeing as much of it as you might have expected by now.

25:46A lot of that is to do with that legacy issue. So it's much harder to put this technology on top of your existing platforms than it is to do it the way we're doing. Now, none of it's easy, but I think that you didn't mention that earlier. I think that's part of the adoption challenge that you're seeing is it's actually very hard to do at scale. You can do it in little bits, but doing it at scale is much harder in a legacy platform world. I think the second biggest challenge that we haven't mentioned is making sure that in two or three years, what we have built is not legacy. Because we always thought like our platform, OK, we've got 10 years before.

26:24If we don't change anything, we've got 10 years before it becomes legacy. Now, with the advance in technology, I don't even know if it's two years, one year, one week or a day or an hour. It's incredible. I always describe it like kind of what Andy was saying is we're running to that full speed, There were these amazing fields that look like paradise, where everything's green and there are butterflies. But we turn around, and basically, it's a road collapsing behind us. And if you stop, you're basically gone. And it really feels like this at the moment. It's both incredibly exciting, but at the same time, quite nerve-wracking.

27:00Because you're like, okay, what are we doing? Are we going in the right direction? Are we doing the things that mean it's still going to be a great solution in two or three years? and we think the way we've built it and because also there was no legacy to build on this project we think everything's been done so it's flexible enough and you can keep on improving and keep on progressing but let's do another podcast in two or three years robin you'll have to come back from your retirement in the caribbean and you can ask us that question we're gonna find out we were keen to get you together to tell us all about kinetic and you have it's a very impressive story i wish you all the best of luck.

27:41And as Guillaume says, I think we would be remiss if we didn't get you all back in a year or two years to tell us how you're getting on, because there's lots of people watching carefully over what you're doing to see whether or not you make a success of it. Good luck. And thank you so much for joining me. Thanks, Robin. Thanks, Robin. Thanks for having us, Robin. Appreciate it.

28:07well if you've made it this far then i'm pretty sure you found that as interesting as i did the instech podcast comes out every sunday morning where we spotlight the latest news leading voices and freshest updates across insurance that you need to know about if you would like to take part in these conversations head to www.instech.co to find out how you can join our network and be a part of the insurance intelligence for the curious.