How Technology Is Transforming Insurance with John Wingate

0:00 Hello and welcome to beyond the Desk, the podcast, where I sit down with some of the most inspiring leaders across the global insurance sector. I'm Mark Thomas and each week I'll be bringing you a senior leader from across the insurance world to hear their story and views on the future of our industry. Before we get into the next episode, I'd be extremely grateful if you could like and subscribe to the podcast wherever you're listening to it, and if you enjoy it, share it with a colleague or a friend.

0:26 It really does help us grow the podcast and get much better guests. Now, without further ado, let's get into the next episode of beyond the Desk. John, welcome to the podcast. How you doing?

0:40 Doing very well, thank you. It's good to be here, Mark.

0:42 Yeah. Well, great to have you. So, look, as always with the these things, I'm going to go right back to the start and kind of go through the journey of your career, but why don't we kind of start as a very brief kind of intro into kind of who you are right now, like what the role you're doing currently, and then we'll go back and work our way back to. To the current role and all that kind of stuff.

1:03 Sure. So, yeah, I'm currently CTO at Envelope Risk. Yeah. Envelope Risk are a cyber reinsurance mga, largely. So we're quite unusual, I think. I think we're probably in a position of one company in that area at the moment. I've been with Envelope for about seven years now.

1:23 Okay.

1:24 So my role is primarily looking after our technology platform.

1:29 Yeah.

1:29 We're somewhat unusual in industry in that we develop most of our own technology.

1:35 Yeah.

1:35 In particular, our technology support underwriting. Yeah. So, you know, we have a very large for the industry in house tech team, which is 35, 40 people now. And, you know, I lead that team. It's my job to essentially curate the content on that platform, to support what we need to do from an internal underwriting perspective and also to support, as we expand into other classes of business and beyond, monetizing some of what we've developed internally to the external market.

2:09 Yeah, amazing, right? We're definitely going to get into that because there's some really interesting stuff there. But let's go right back to the start of your career because obviously, I know, I'm pretty sure Envelopes are for your first insurance role, but you've been in it for quite a while now. So. So what did that. What kind of did the early days for. For John look like? Like, were you. Were you Always into kind of technology and stuff as a kid and like, I mean, what, what did, what did kind of early years look like?

2:32 Yeah, I mean, I think, you know, I've always been interested in technology and I've always been interested in maths.

2:39 Yeah. Okay.

2:40 Always fascinated me that you could kind of describe the world using equations. Yeah, describe and predict the world. That, that, that always fascinated me.

2:49 Yeah.

2:51 And, you know, I grew up in the generation when, when kids first started getting computers in their bedroom.

2:56 Yeah.

2:58 And rather than playing games, I like to build my own games and create my own stuff.

3:02 Yeah.

3:03 So that, that, that really got me and this sort of blend in maths and computing has, has always fascinated and intrigued me and I, I guess that led on to, to what I studied.

3:15 Yeah.

3:15 At university.

3:16 What was that?

3:17 So I studied aeronautical engineering.

3:19 Okay.

3:19 At University of Glasgow, which was a kind of blend of applied maths and computing to airplanes, which I was fascinated with as well. So. And you know, that was, that was great. I loved doing that. It was a fantastic, fantastic subject to study. A great time when the world was changing from digital perspective.

3:44 Yeah.

3:45 And you know, being at this all Bolton floor, that was, was very exciting.

3:48 Yeah. Where did, and where did you, did you grow up in Glasgow? Is that where you live?

3:51 Yeah, I grew up, born and brought up in Glasgow.

3:53 Yeah.

3:53 Yeah.

3:54 So there's still a bit of an accent there. It's not probably quite as strong as it maybe was.

3:58 Yeah, I've been sort of, you know, 20 years in the west country now and I still still managed to keep the accent. Yeah, yeah, but, but, yeah, so, yeah, I went to my local university, which is quite common in Scotland. A lot of people go to there.

4:11 Yeah.

4:11 Local universities. The, the way the, the school system works in Scotland is if you do good in your exams, you can actually start a year early.

4:20 okay.

4:20 So I started at 17, which, you know, it made more sense to, to be local in that case.

4:26 Yeah, of course. So she did that for a few years and, and then. So what was the first role out of, out of university? Did you, did you kind of stay on the, the theme of the aerospace type stuff?

4:39 Yeah, I was, I was very lucky. I, I got a job in, you know, working jet engines.

4:44 Oh, wow. Okay.

4:45 So. So the early part of my career was, was really working, developing software for controlling and managing jet engines.

4:53 Yeah. Was that, was that kind of what you wanted to do? Like, I mean, was that, was there a plan out of university?

4:58 Absolutely, absolutely. In fact, you know, ironically, I'm in, I'm in Insurance. Now I kind of had two job offers really. One, you know, to work in jet engines. The other one was an actuary.

5:08 Oh, wow. Makes sense.

5:10 By a roundabout route again. So yeah, it was a great place to be. I was able to not work just on the software, but working on developing a lot of the modeling, mathematical modeling behind those, trying to predict what the engines were going to do, design control systems around them and going back to describing how the world works with equations. I always remember earlier on in my career spending six months developing this very complicated mathematical model of an engine, designing all this software to control it and then getting on a real test bed and pressing the button and the thing started and control.

5:56 It's like, wow, this stuff.

5:57 Yeah. Doing some action. Yeah, real. Yeah, yeah, yeah, yeah. So, and so who were you working for doing that?

6:04 So Rolls Royce.

6:05 All right, okay, cool. They're like the main player in that space.

6:07 Yeah, absolutely. So I, that's how I ended up in Bristol. So yeah, of course I relocated to Bristol to work with Rolls Royce.

6:15 So you. So you relocated down to Bristol quite a young age and then you're kind of early 20s.

6:19 Yeah, yeah, yeah, yeah. So I've been there for quite a while. I, I did a bit of a stunt in Birmingham beforehand.

6:26 Yeah.

6:27 But you know, ended up in Bristol and you know, started off jet engines, you know, aero jet engines, and then did a stint in marine jet engines.

6:36 Right.

6:37 You know, marine gas turbines. So did a lot of work on the engine that actually powers the, the Queen Elizabeth aircraft carriers.

6:44 Yeah. Okay, cool.

6:45 I led all the software development.

6:47 So it was all military stuff, was it kind of defense.

6:50 Defense and civil also. And you know, again, the kind of emphasis changed from being directly just looking at control of the engines, but to sort of look at the wider loop around how to do predictive maintenance and, you know, minimizing ownership costs and dealing with all the uncertainty around that. And really it was all about developing systems that make. Allow the experts to make better decisions under uncertainty.

7:20 Yeah, yeah.

7:21 Which sounds a lot like insurance.

7:23 Yeah, yeah, yeah. So you can see the synergies. So, so that was kind of very much kind of hands on engineering type stuff. What was it the evolution look like there? And you mean how long did you spend at Rolls Royce?

7:35 So in total at Rolls Royce it was probably north 10 years.

7:42 Okay. So quite a long stem, quite a long same role or kind of evolving.

7:47 Evolving. So, you know, started off as a graduate trainee and then, you know, I was leading software teams and systems engineering teams.

7:54 Yeah.

7:55 Towards the end, you know, the MT30, the marine tread. So I was leading this software development on that. It's a great place to learn that skill because you pretty much figure out pretty quickly when you've got stuff wrong.

8:08 Yeah, yeah.

8:09 Things explode.

8:10 Yeah.

8:12 And it's one of the key differences with insurance relative to some of the other areas I've worked in with. If you're working in real physical machinery and real physical plant, if you get your software wrong, you'll find out in 30 milliseconds, whereas it can obviously take years. Insurance.

8:31 Yeah, yeah. And I guess there's got a kind of mission critical nature to them as well.

8:35 Yes. Different level.

8:37 Yeah, yeah, yeah. So, so, so what was the kind of catalyst for you to move on from there?

8:42 So I, I, I, I'd been there for a while. I wanted to do something new. Yeah, different. I've been getting more involved in kind of the energy element of the business.

8:53 Okay.

8:54 And you know, an opportunity, opportunity to come out to work with General Electric.

8:59 Yep.

8:59 Okay. Another huge company in their oil and gas division.

9:03 Right.

9:05 Did a lot of work on subsea production systems. Again very similar systems around predictive maintenance, figuring out when stuff was going to break down so you could optimize all the logistics of repair, et cetera around that. I then got the opportunity to join GE Software, which was GE wanted to hop on the software and eventually made huge investments, you know, set up huge offices in, in the Bay Area.

9:38 So you know, I worked there for a number of years and relocated over there. I did a lot of traveling.

9:43 Yeah.

9:44 You know, working on a whole host of real world problems. And this, this was really at the time where, you know, the, the machine learning, the AI bubble was, was really kicking off. So I had the opportunity of working in the Bay Area, applying this to real problems with big teams with huge budgets. So it's very exciting times. Yeah.

10:04 So what, so what kind of level in the are you at this point? Have you moved into pure management at this stage?

10:10 No, I mean I've always stayed hands on. I've always liked to stay on top of the technology and roll my sleeves up and code and do the modeling, et cetera. Ge, I go up to analytics director.

10:27 Right.

10:28 So for G O and Gas. So you know, we were working with big clients all around the world.

10:34 Yeah.

10:36 It got to the point where I just got sick of the traveling.

10:40 Yeah. Novelty wears off after all.

10:43 Cost the market. One of my major customers were in Australia. So managing time zones between California, Australia and the UK was tough. Yeah, was, was tough. I had young kids at the time. Yeah. So it was time to look for

10:57 something new and there was the next role, the envelope one.

11:00 I had one before that.

11:02 Okay.

11:02 So I joined a startup machine learning consultancy in Bristol. You know, again, it was a great opportunity, a great company. We were really pushing the boundaries and applying machine learning in the real world. I had the opportunity to build a really good team there. You know, we were working with some big clients on big projects and we were doing stuff from, you know, applying ML to planning in London through to, you know, electricity transmission in San Diego.

11:39 So it was a whole breakfast, a

11:41 whole bunch of stuff and really learned how to, you know, engineer this technology and build the teams around it. Yeah.

11:49 So was that, was that a long stint there, doing that one or so?

11:52 Yeah, I mean I was there for.

11:53 Were you one of the kind of founding people in that. In that. Yeah, so.

11:57 So the, the bits had kind of been going, but they're more of a sort of web app development company. But I joined and we sort of pivoted.

12:04 Yeah.

12:04 Into, into the machine learning area. So yeah, I would say for about four years.

12:10 Okay, so it's a decent stint.

12:11 Yeah, yeah. You know, built up pretty good team there and you know, had some, some good successes.

12:17 Yeah. So what made you leave there to go to Envelop?

12:20 You know, I think it wasn't panning out exactly the way I wanted it to do. And you know, one day I got speculative call from, from a headhunter, you know, and, and you know, they described this, this rule in reinsurance, you know, cyber insurance. Yeah, A business which wanted from the get go to build a tech stack on ML and AI. And they were looking for someone with a real sort of applied background like myself to join the team.

12:55 I spoke to the guys and thought, yeah, this sounds fantastic. So I joined.

13:01 So did you join as cto?

13:03 I joined initially as head of data science.

13:07 Okay.

13:09 So there was. One of the founders was there and within a year or so we kind of swapped roles.

13:18 Oh, so he was the cto, but kind of found he had the kind of founding idea. Okay, cool.

13:22 Yeah, so it made sense. I had a lot more experience of hands on IT switch for building larger teams, applied technology and scaling things out. Yeah.

13:36 Okay, so let's get a little bit into the invalid one because I mean the, the, not everyone will necessarily know the business. Not, not necessarily. Kind of a huge insurance name like another one of the big carriers or something. So, so I guess let's, let's hear the kind of the pitch of the business. What, what, what do you guys do. And, and then we can get into a little bit more about what your, your role has been over the kind of seven years.

14:01 I know you've done a lot of stuff there.

14:03 Yeah. So, I mean, Enveloped Risk presently are

14:08 a cyber reinsurer purely, Purely in cyber.

14:12 Purely in cyber. But we are expanding as we speak across other classes of business.

14:17 Okay.

14:19 We, you know, operate as an mga, but we also have skin in the game. We have our own capital vehicle. So we sort of co invest.

14:26 Yep.

14:27 With our capacity partners there. I think our value proposition is really to provide our capacity partners with better return on capital relative to the rest of the market. The technology that we've developed and the way we work with our underwriting teams really allow us to build those and engineer those capitally attractive portfolios. Yeah. In essence, that's, that's, that's what we do.

14:59 Yeah.

15:00 And so is that purely a tech play or is it, I guess you still have kind of teams of underwriters.

15:05 We still have, yeah, absolutely. You know, we, you know, we make our money from underwriting.

15:11 Yeah.

15:13 You know, with, you know, commission and. But we've also got our own capital vehicle, so we have some risk capital in there. For myself also. Yeah, yeah, yeah, yeah.

15:23 It's just underpinned by a heavy tech tech element and I guess what's the, what's the, the USP from a, from a tech perspective, what are you guys doing that, that nobody else is, is, is doing?

15:36 You know, I think, I think we have, I think revolutionized the way that technology is applied in this area. Right. So, you know, we've written get on one and a half billion premium since we started. We do that with a very small underwriting team. Very highly skilled. A small underwriting team. Much of the sort of technical underwriting element is actually carried out by my team.

16:08 Tech team. And the main process is within there. I think you can break those down into three areas. One is the ingestion of data. We've built technology that allows us to automate ingestion of very complicated, diverse border rows. This can, you know, in other companies can be a process which can take weeks to do. We can turn that around in minutes. Wow.

16:37 Okay.

16:39 We have built a very sophisticated cyber risk model. You know, we've been able to do that because, you know, one of the key strategic reasons we set up as a reinsurer was it gave us an access to pretty much all of the industry's data.

16:53 Yeah.

16:53 And as a machine learning and AI shop. That's the fuel. The fuel's what we do.

16:57 Yeah, yeah, okay, makes sense.

16:58 So we were able to build a very sophisticated cyber risk model. A cyber risk model that was really targeted to allow us to write non proportional business and gave us the real insight that we needed to do that rather than just buying a third party model. And there's some great third party models out there. And the third I think most important layer that sits in that is the decision layer that our underwriters use. So you've got all this data from Bordero describing what your exposure is.

17:33 You have a very fine grain viewer risk. How do you turn that into a decision which is what our underwriters need to do? We built a layer that sits on top of that which will effectively allow our underwriters to plan and follow that risk. From planning all the way through to risk runoff. It allows a very detailed planning process. And as a result of that planning process, we can provide optimal participation in each risk that we want to underwrite to maximize our underwriting profit within our underwriting constraints and our capacity providers underwriting constraints.

18:18 So this is all done in a very big, very sophisticated optimization loop which is really capitalizing on some of the commoditization of hardware that's happened. New age of AI GPUs are now cheap, very large scale efficient optimizers. That's what we're using. And we're giving our underwriters recommendations on line sizes. The market may speak, they may not get that line size.

18:48 That's fine. We hard code what we got and then squeeze as much juice as we can going forward for the rest of the year. So it's this virtuous cycle and I think we have unique in operating in that way and it's allowed us to gather huge amounts of data and really leverage that. Not just an abstract quarterly, once a year exposure roll ups and models, but in day to day underwriting operations.

19:19 So what does that look like for kind of. You mentioned you spanning that out into other classes of business? Is it just kind of applying the same type of technology in businesses, business lines that are kind of complementary to that? Or is it kind of a model and approach that you can in theory span out to any other line of business?

19:43 In theory, yes. I mean the delta between some is less.

19:48 I guess there's some that are more complex than others but.

19:50 But essentially two of the main components here, the border or ingestion for instance, is pretty much like an agnostic. The optimization capital management layer just described is independent. We have a very, very sophisticated view of risk in cyber, but we can cold start in other lines with la less sophisticated models. And once we start the flywheel going of getting in the data, we can then, you know, look to, you know, approach what we have in cyber.

20:25 Now the underlying architecture of the cyber model lends itself to, you know, any other line of insurance, really.

20:33 Yeah. Okay. And so your role there. Let's, let's talk a little bit about what the, what you kind of started doing and what that's kind of evolved over the last kind of seven years. Is it, is it been kind of a, a growth thing in regards to kind of building out that team, doing more capability? Have you, are you doing very different stuff now to what you were doing when you started?

20:55 Yeah, I mean, I think, I think it has been a logical extension.

20:59 Yeah.

20:59 Of where we started from. I mean, when we started, when I joined Velop, we just set up our office in Bristol. You know, there was three or four of us in the company. You know, very much focused on that cyber risk quantification element. You know, as our books grown, then you know, the need to get the inflow of data into the system more rapidly, that was the big bottleneck. So we've invested a lot in doing that and doing that is accurately and as quickly as we can because, you know, being first, to quote, is, you know, a real important differentiator from us relative to the rest of the market.

21:48 So, you know, we started, we've got better at it, we've got more efficient at it. Our book has been growing over time. We now, as I said in my team, do much of the technical element of underwriting. We have developed in house tool to make that more and more automated over time and to make our small underwriting team as effective as possible on putting the tools in their hands that make their decision making and their evaluation of the risk easier.

22:25 And you know, they can concentrate on some of the softer, more human centric aspects of the job. Although they're still very involved technically. I think our technology has been a great recruitment tool for getting underwriters into the business.

22:44 Yeah. And so where does that go as an evolution of your role? Like kind of the, if you look at the future, is the main kind of task for you now, applying that technology to the different lines of business that you choose to go going, and what's kind of determining where that goes into the kind of next direction?

23:06 Yeah, I mean, I think that's part of it. And you know, there's. How we address going into those other lines of business will dictate that. And you Know, we are, you know, we have announced probably a year and a half ago or so the envelope solutions business. We were looking to make some of what we've developed internally available to key partners within our ecosystem.

23:36 So we're moving beyond this being purely an in house tool to get into the stage where we can share and use it with external partners.

23:45 Yeah. And so is that a model that kind of other carriers and other MGAs might kind of plug into or is it kind of subscription type thing? How will that work?

23:55 You know, I think there's lots of options on the table about how we do that, but essentially, you know, whether it's a license cost, whether it's some percentage of gwp, whether it's a profit share in there, we're very flexible about how we do this. Yeah.

24:13 So that's the next kind of phase of growth for you, is kind of building out the technology team to support that, building that, the solutions business. I guess that will turn you into more of a, a kind of fully fledged technology business rather than just purely an mga.

24:29 Yeah, I mean, I think, you know, I would argue we've always been a technology business.

24:33 Yeah, at its core.

24:34 Yeah, at its core. I mean that's, that, that's kind of been the founding principle of the company. You know, I think, I think my role, you know, becomes much more of, of platform and product as well as, you know, technical direction. Technical direction. You know, understanding what commercially makes sense for us to introduce into the platform. You know, we have this sort of analogy of, you know, one kitchen and many, many restaurants.

25:06 So you know, understanding what goes on the fixed menu, what goes on the daily specials and what can be done as a special. So you know that, that's very much part of my job is my, my, my rug rules to make sure that's all done in an efficient manner.

25:20 Yeah. And, and how did you just kind of, I guess stepping back to kind of pre Envelope now? Like, I mean, I mean I'm always intrigued at how people found the kind of transition from, because of what you were doing before was very different from an industry perspective. But it doesn't sound like actually the, the leap was quite as, as drastic is what it would have been for someone that go, goes from working kind of in that to working for one of the big carriers at a bit more kind of traditional insurance.

25:48 Because actually it sounds like Envelope was probably doing a lot of very similar stuff that you'd done before, but just applying it to a different industry. So was that, was that jump a drastic one? And like, I mean how did, how,

26:02 how did you find that from a technology perspective? Not really. Yeah, I think, you know, the, the biggest jump was just getting to grips with all the industry jargon.

26:11 Yeah, understanding it.

26:13 Yeah, that, that was, you know, especially in the world of reinsurance, you know, there's not much documentation around for someone to, to, to come up that curve very quickly. But luckily, you know, envelope we've got some real industry experts and veterans there.

26:29 Right.

26:30 Where, where you know, had the opportunity to up that curve relatively quickly. Yeah. And you know, we, we didn't, you know, when I joined we, we didn't have the, you know, we weren't sitting back designing this stuff. We were doing it for real while we were underwriting. So you had to learn how to do it pretty quickly. Yeah, but you know, in terms of underlying technology that we are using, I mean, you know, I, I developed my first neural network in the 90s for aircraft flight controls.

27:05 So it's been a steady evolution. Time scales are maybe a bit different, there's bigger errors in the measurements, there's perhaps a bigger human factor in there, but essentially the same frameworks still apply I think perhaps. And I think this defender and were probably onto something here in as much as wanted to bring some of that thinking into the world of insurance.

27:44 Was it always their business plan to do what they've ultimately done with Envelop or did that kind of evolve over time?

27:55 Yeah, I think there was some ambiguity in the early days but essentially the founding team at Envelop was Jonathan, who's sort of an insurance financial services veteran and Ari, who's a cyber underwriter. And Paul and John were both from the tech world, Lockheed Martin and NASA. So they came with engineering. Yeah, your defense aerospace background as well.

28:26 So there's a bit of a connection there with you and the kind of backgrounds you've done actually, which made sense. Yeah. A quick word from the sponsors of beyond the Desk, my business, Invector Group. Invector is a PE backed specialist executive search firm working exclusively in the insurance sector. We've supported over 25 of the world's leading insurance and insuretech businesses to identify and secure key leadership talent across actuarial data, AI technology broken and underwriting. Our team of deep industry specialists typically deliver searches in under four weeks and we consistently outperform traditional search firms on speed, quality and candidate fit.

29:06 So if you're looking to make a business critical or senior leadership hire and want a modern specialist who genuinely knows the market in detail, we'd love to Talk to you for more information, connect with me on LinkedIn or visit us@invector group.com now, let's get back to the episode. So you mean you've touched on it a little bit kind of the, the AI side of things. I mean it's obviously hot topic now, although it's sounds like you were kind of playing around with this stuff way before it was kind of the top of everyone's news agenda.

29:37 So what, what's your, your kind of as someone who's been in the. That industry and been playing around with stuff like that for. And, and, and, and you mean it sounds like kind of just using it in, in real life stuff for quite a long time. What, what on where we're at currently and what the next kind of few years look like from I guess from a world perspective, but also an insurance perspective as well.

30:03 Yeah, I mean, I think there has been an acceleration in the capabilities of AI, I guess driven by Frontier models, your open AIs and etc. I mean these things are great, super cool. I've been watching them evolve. I think one of the. I think OpenAI were very clever around how they introduced this.

30:35 It was introduced in a very usable way.

30:39 What is in the kind of chatbot thing?

30:41 The chatbot thing where people could for free actually see this.

30:44 Yeah, it was just kind of, it came out of nowhere really.

30:46 But if you'd been working in it, you'd see.

30:48 Seen these things. Yeah, yeah. I guess for kind of Joe Public, who's not technical like someone like myself, it just seemed to kind of just appear and then people were using it. I don't think they expected it to be, I mean the general consensus, they didn't expect it to be as good or as heavily used as what it actually ended up being.

31:03 Absolutely, absolutely. And I think, you know that that underlying frontier model will continue to improve. But you know, a lot of the, I think innovation around it is how you engineer systems around this very clever frontier model which is generating text. And you're seeing the likes of Anthropic, their coding tools are absolutely amazing. OpenAI codecs as well.

31:38 This is, as we speak, revolutionizing coding and software development. And it's not surprising that this is one of the areas where this has shown the most promise. First, because you can measure how good software is with unit tests and bugs and all the rest of it so you can iterate really quickly and improve how it generates it. And that's genuinely impacting the way my team work today.

32:12 Does it mean that we are Going to be recruiting and laying off engineers? No, it just means we're going to write more software. So projects that 18 months ago, two years, we wouldn't even start doing. Because I want to recruit 20 front end developers.

32:31 We are now because it might take one or two to just kind of oversee it.

32:35 Yeah. So, you know, that's a huge difference. So we're going to be writing more software and that applies across a whole range of technical tasks. And you know, we can do more, we can do more quickly and we can do it better. And you know, I think, you know, I'd be lying to say this was via design, but you know, the kind of team structure and culture we have at Envelop is all around really immersing yourself in the domain and that blend of domain expertise and having our arms around how to use this technology gives us a real advantage.

33:17 And so is the. It's interesting that, I mean, I've got so many questions about AI that I'd love to kind of get into a bit because just for someone who knows more about it than, than most I would imagine, I mean, are you seeing that it, it's kind of, is it an advantage the fact that you guys understand this significantly better than, than the majority of people? Because on the face of it, you'd think that that is a good thing.

33:44 But you guys were doing cool stuff with this before everybody else was looking at it because they, it was almost kind of forced upon them in the last couple of years that it's like there's a mad rush to try and use it more effectively. or is it more of a, kind of a negative now that other people are trying to catch up? Like, do you see what I mean?

34:01 Yeah, I mean, yeah, I mean, I think, I think, you know, there's always two ways to look at things.

34:06 Yeah, yeah.

34:08 You know, it's very, very easy to build the wrong thing with this technology. Yeah, exactly.

34:14 Because it's easy to build and quickly and you can do it quickly.

34:16 And you know, a lot of the skill in putting these systems together is really understanding the problems, understanding the right problems to solve.

34:25 Yeah.

34:26 And it's, it's, it's more about data and workflow. So it's understanding where you need to inject this and how people operate day to day. And I think by virtue of the way we've operated and other businesses in the same way we, we have an unfair advantage there.

34:45 Yeah, yeah. Has it, has it changed the way that the, the types of people you hire now? Like, and I mean I, I don't necessarily mean it. I mean, possibly in technology, but, but in the wider business, are you, are you seeing a kind of a shift in, in the kind of people. Human capital strategy because of AI as yet? Because I think that's quite an interesting thing. Like, are people now hiring people that, that have to be AI savvy or were you guys already doing that?

35:15 Yeah, I think, I think we were probably somewhat ahead of the curve.

35:18 Yeah.

35:18 Because we, you know, technology was always do. And, and you know, I think, I think we attracted people who were interested in things.

35:28 So you don't. The lack of adoption or kind of openness to that kind of thing is probably. The bar is a lot lower.

35:34 So we were, we were very much seeing a biased sample.

35:37 Yeah.

35:37 Of the market.

35:38 Yeah. Yeah. And what about kind of more generally in insurance? Like, you mean, do you think, you think the industry as a whole is, is adopting AI in the right way? From, like, given that you guys have been kind of ahead of the vast majority, like, are you seeing any kind of interesting use cases or do you think it's all.

36:01 Yeah, I think, I think there's this. There's some pockets of really cool stuff being done. You know, I think like many industries, and this won't be unique to insurance, lots of industries which are regulated and, you know, have, have a lot of inertia in their own internal processes will see the adoption of this technology more. More difficult. And I think we will see some inertia there.

36:33 But ultimately I think the benefits will, will overcome that. And I think as with the introduction of any new technology, some of the easier, more manual work, which is more easily automated, will be automated, but the skill level will move up. And one of the advantages of this technology is it means that it can reduce the learning curve for people to move up.

37:09 Yeah.

37:09 That skill level. So I think that's, that's what we'll see.

37:12 Yeah.

37:14 And I, I'm not quite sure how, you know, I've heard this, this example thrown around several times. I'm not quite sure exactly how, but it's a good example. It's, you know, one of the areas that were, you know, seem to be victim of machine learning and, you know, pattern recognition was, you know, radiology, radiographers.

37:34 Yeah.

37:34 You could get ML to look at these scans and pull this information out. But there's more radiologists now than there ever was.

37:41 Yeah, I heard that the other day.

37:42 Yeah. So people are just doing more.

37:44 Yeah.

37:45 Of it. They're doing more scans and you know, where their decision in that human element comes in is just moving up higher level of abstraction. I think that's what will happen in general.

37:54 Yeah. Yeah. I mean I think the example they give about radiologists is. Is that kind of you. You still need people to. To kind of set treatments for people who have got things wrong with them. They. You still need that kind of bedside manner. You still like. I mean you're not gonna. We're not quite at the point where someone that can robot's gonna walk into a hospital bed and start kind of talking to it to think it's certainly not yet anyway.

38:17 And I mean I'm not sure you ever really get to that. Like does anyone really want that?

38:21 Yeah, yeah, yeah. It's.

38:23 It's an interesting one. So do you. Do you see I guess more kind of broadly speaking are you kind of optimistic? Like I mean there's, I mean I'm. I quite. I listen to quite a lot of podcasts. I try and keep it fairly rounded but there's you. You can. You can listen to one on a Monday and, and it be super optimistic about how we're going to live in some kind of utopia world. You listen to one the next day and it's.

38:47 It's. It's the total opposite. A kind of fairly doomsday type thing. What, what's your. Your general kind of view on. On. On what the future looks like or. And certainly in the next few years.

38:57 Yeah. I mean by nature I'm a d. Scottish engineer so you, you would. You'd expect me to be pessimistic about it, but I'm actually very optimistic. Yeah. I think I, I think it will revolutionize and she's who new classes of business that people haven't even thought about yet. Yeah. And in the medium to longer term.

39:20 Yeah.

39:23 As have all technologies from you know, the industrial revolution or not. It's been the general trend.

39:28 Yeah.

39:30 And you know the, the businesses who are the biggest businesses in the world in 20 years time. We probably don't even know what they are yet.

39:37 Yeah.

39:38 You know they will definitely be there. Yeah.

39:41 Yeah. I mean I think there definitely seems to be at the moment it's, it's more actually just being able to do more rather than actually kind of let's. And, and if the. I mean I can't remember the exact stat but the, the kind of the numbers around what the pace of change looks like kind of years change in like a week or something like that in. In however many years that that just means you're doing more really doesn't it?

40:05 Yeah, more could doing more and doing more, doing stuff more quickly. So whether that means less humans or just kind of robots doing the, the, the, the difficult stuff and, and being able to, to do more, have more output, I don't think you do that without human. So, yeah, I mean, I've gone round in circles a little bit depending on which who I've listened to that week, but certainly it seems at the moment it's.

40:30 Yeah. I mean there's a really interesting carving. I can't remember the name of it, but it's a measure of human progress from the Industrial Revolution. So it's an exponential curve, it's sort of going up like that. So it's a straight line sheet process in a certain way. And in order to stay on that line, which we've been on since the Industrial Revolution, you have to have accelerating progress. So you're always going to see progress accelerating to stay on the line.

40:59 And there's a lot of very clever economists and observers of this thing have been watching this line very carefully since the latest onset of AI. And it's not shifting.

41:11 Yeah, yeah.

41:12 You expect to see the acceleration.

41:15 Yeah. And there's normally a correction, isn't there like the end. I mean, I think a lot of the, the, the kind of chatter around at the moment is, is that the amount of money that's just being invested into AI can never kind of, not everyone can win. Right. You know, everyone. There's going to be some significant losers out the back of it. And, and yeah, there will be a bit of a correction at some point, but that, that's, that's fairly normal.

41:37 Right.

41:38 It kind of hockey sticks up, there's a, there's a drop and then, and then, and then it kind of carries on on a slightly more steady note when it, when it corrects. But yeah, I mean, it's not, it's not going anywhere, is it?

41:49 Yeah, the dot com bubble.

41:52 Yeah, exactly.

41:53 Yeah. But yeah, I mean, you know, and this isn't to underplay, I mean this, you know, the, the technology and impacts that this has will be huge. You know, and you know, I think, I mean, there's been some announcements recently around this new anthropic Mythos glass wing.

42:15 Yeah.

42:16 You know, uncovering all these vulnerabilities and you know, what's the impact of that? And you know, you hear speculation on one side is, you know, it's going to be the end of the world, it's going to be cyber apocalypse. Yeah. And then the other side of it, you hear well, you know, using this technology, all vulnerabilities will be eliminated and there'll be no need for insurance anymore.

42:37 Yeah, yeah.

42:37 So you've got both these counts and as always, the answer is going to be somewhere.

42:41 Yeah, yeah, yeah. And then I guess, I guess more kind of broadly speaking in, in the insurance space. And, and what, what do you see as the, the kind of, the big areas of kind of for trends over, over the coming years where you, where you think people will, where there'll be changes in the, in the industry kind of technology, that kind of thing.

43:03 Yeah, I mean, I think, you know, the ones that will be addressed first are those really document manually intensive workflows, you know, to document ingestion or ingestion, claims handling. You know, those are, those will be hit first and we've seen big productivity improvements there, you know, in the sort of medium term. I see, you know, insurance products at the moment are fairly generic.

43:37 You know, they cover a broad range of bases and oftentimes people can't get exactly the COVID they want. I think one of the key side effects of this for the insurance industry is getting much more specific around risks. Completely bespoke, almost personalized risks. So you know, if you look at a company's risk register, here's, you know, you can imagine at some point in the future, here's the risk, here's your risk. You right click on it, you know, how much of that do you want to cover with insurance?

44:12 Very, very specific. And being able to do that very, very fast.

44:15 Yeah. Is that, you mean you think that. Is that because the ability to be able to look at the data and, and therefore come up with a price will be like exponentially quicker. So like at the moment that's impossible to have completely bespoke because it just takes too long every time.

44:33 Absolutely. So, you know, and that goes into policy and handling the claims associated. But you know, once, once we can really weaponize this technology in an insured sense. Yeah. Then, then I think we have an opportunity to write much more targeted insurance and really mitigate the risks that companies and people care about.

44:54 Yeah, yeah, that's an interesting one.

44:55 Yeah.

44:56 Because I mean you can. If you can look at that data that much quicker, then there's not really

45:01 any reason not to, you know. And you know, we think about is, you know, also be able to onboard additional capital into the industry and give them confidence that you're pricing this risk correctly.

45:13 Yeah, I mean kind of hyper personalization is. It seems to be a common theme across lots of different industries. And it like, I mean simple is kind of being able to write emails now that kind of know everything about you and not not have to necessarily do all the research and stuff to.

45:27 Yeah.

45:28 Kind of every, everything you get offered seems to be kind of hyper personalized now. It seems to be the way that the world is going generally speaking. And, and I guess I just wanted to move on to kind of more generally about the, just some kind of the key lessons you've learned in, in your career as well. I mean I'm always quite interested to give kind of listeners a few snippets of gold to take away from them.

45:50 But I mean you, you've obviously been various different roles, different industries like now ended up in, in insurance. But, but I guess a common theme kind of running throughout it. I mean have there been any kind of things that you would pick out through that career journey that were really key learning points or kind of inflection points that you've had throughout your career?

46:13 Yeah, I mean I think, I think there's, you know, being able to communicate clearly.

46:19 Yeah.

46:19 Is super important.

46:21 Yeah yeah.

46:22 And, and communicate at the right level. Yeah. The person that you're communicating with. You know, I think quite early on in my career I sussed it up was very important and invested quite a bit of time into doing it.

46:38 And I guess you also find lots of kind of stereotypical techies that, that, that's, that's often the kind of a lot of the barrier to, to being able to, to kind of move up the, up the ladder, I guess, isn't it so?

46:51 Yeah, absolutely.

46:52 I think it's less so now in fairness than, than probably before, but certainly I experienced that now a little bit less so.

46:58 But. Yeah, yeah, and, and it, it's certainly something, you know, in my teams I try and encourage people to get up front of the team, do presentations, get more confident expressing themselves, get more comfortable having an opinion. I think it's a really important growth area for young engineers and scientists that I work with. I think it's also really important to be able to empathize with other people and see the world through their eyes and really understand what they're optimizing for.

47:34 Yeah. And you know what incentives move everything. You know, what's their incentive and trying to understand the way they see things. That's, that's super important.

47:45 What. Were there any things that from working in the, the kind of the aerospace, military, defense type of, of world that, that have particularly kind of been useful in working in the industry you're in now? That maybe other people who hadn't worked in that, in multiple industries would, that you've kind of got in your toolkit that maybe they didn't, they wouldn't have.

48:11 Yeah. I mean, I think, you know, the industries that you just mentioned are all very much characterized by, you know, the underlying concept behind everything is safety. Okay. And always thinking 10 steps ahead of everything that can go wrong.

48:29 Right.

48:30 Is super important.

48:31 Yeah, yeah. Suppose you've got kind of lives at risk and stuff every day.

48:34 Right? Yeah, yeah, absolutely. So I think that can, that, that's, you know, perhaps it's, it's, it's not the best way to look at the world in some cases, but we really always stress testing everything to see what could possibly go wrong.

48:52 Yeah.

48:52 And making sure that you plan to mitigate that where possible. I think that really characterizes the thought pattern in those industries is always safety first.

49:05 Yeah. And what about the kind of, any of the kind of big wins or the big decisions you've made in your career? Is there anyone that kind of stands out as the, as a kind of a point that you made? You had to make a kind of 50, 50 call. You mean maybe the defense versus actuary that ended up coming back full circle?

49:31 I, I'm not, I'm not sure. There's been like one big.

49:35 Yeah.

49:35 Massive master decision that's, that's changed everything.

49:39 Yeah.

49:41 I think, I think in general I've been quite lucky in that the career decisions that I've made. Yeah. Have been the right ones at the right time.

49:48 Yeah. Yeah.

49:49 You know, I think, you know, looking back in retrospect, and I won't mention the companies, but understanding when the right time to leave is often important.

49:58 Yeah, yeah.

50:00 And you're being thoughtful around where your next role is going to be. Not just in terms of looking at immediate monetary, you know, advantage, but thinking about, you know, the way the economy's going, the way that industry is going.

50:17 Yeah.

50:18 You know what, one of, one of the reasons I found going into insurance and cyber insurance so attractive was just thinking about, you know, how the world works, you know, going back to, I don't know, 1980 or something like that. 80% of the world's capital was tied up in physical things. And you know, we had a very mature insurance industry to help mitigate that risk. You know, fast forward to now, it's flipped. You know, 80, 90% is in digital non tangible assets.

50:50 Yeah. And you know, you know, that, that, that, you know, the risk mitigation insurance right now is still relatively early days so it's a huge opportunity. Yeah.

51:00 I mean, did you always have the, did you always know when you started that you kind of had ambitions to go to be a move on to be kind of a cto key exec leadership type role? Because I think that's the, I mean I, I've said it a lot on this podcast. I mean I talk to a lots of aspiring CTOs, CIOs, CDOs, and, and they've typically come from a technical background. Not, not always, but, but a lot of the time they have.

51:28 And, and certainly I would say there's just as many that, that make it to that role that, that enjoy that, that don't enjoy it because they miss the technical part of it and stay in hands on. I know you said that's kind of been important to you to kind of stay in the, in, in the detail, I guess in, in a role which is kind of still fairly technical, running techno technical teams, maybe slightly different to kind of the, the, the CIO of the, the big insurance carrier or something like that.

51:57 But, but yeah. Did you, did you, did you always see that as your path and, and, and, and, and have you ever, has ever been that point where you've kind of missed in being really entrenched in the detail?

52:10 Yeah, I mean, I think I've never deliberately sort of pursued being a career manager. I like what motivates me is working on large important projects that make a difference and you can't do that yourself. So in order to do that you have to, you know, have a team. And I've always seen a team as being an amplifier to allow me to do that.

52:48 I think, I think that, I think it also makes you think about the team differently as well. So it's, it's, I mean, to me the big thing about, you know, you know, having having a great team around is really understand how to trust them and you know, make sure that, you know, you have to be very careful around recruitment and get people who are just sort of laser focused on the very technical aspects. They've got that sort of system thinking aspect to them and you want to be able to trust them to do stuff to, you know, help meet your goals of that bigger project.

53:26 I think, you know, as I've, I've progressed in my career, I've done bigger and bigger projects and had to bring a team on board to, to do that.

53:36 Is that been the main, it certainly sounds like it. But has that been the main driver for you when you kind of. Because you've had You've had kind of four key businesses that you've worked in. It sounds like the, the main kind of catalyst for you choosing those companies was interesting stuff to do rather than necessarily a kind of a job title or. Yeah. Like a kind of big, big role or anything like that. It sounds like it's always been when you were talking through that, that career journey, you were always talking about what you were doing rather than the role that you were doing, so.

54:11 Yeah, yeah, absolutely. I think, you know, that's, that's what gets you to bed in the morning.

54:17 Yeah, yeah, yeah. And then I guess the, the I'd be really interested to know what your, your view on what makes a good Chief Technology Officer in like and if you've got what, what would you kind of give advice to someone who wants to, wants to move to that point. What do you think makes a, a great cto?

54:40 I think, and I'll give you how, how I've done it. Yeah. That's the best CTO or not. But you know, stay technical.

54:52 Yeah.

54:52 You know, obviously you can't sit down and do technical work every day but, but absolutely stay abreast.

54:57 Do you think that's really important? Because I think a lot of people move away from that, don't they?

55:00 No, I, I think, I think it's important. I think it's, it allows you to make better decisions. Allows you to gain respect to the team.

55:08 Yeah.

55:09 It allows you to understand the trade offs better. There's never a silver bullet solution to any problem in the real world. You get a better handle on how to understand the trade offs better. I said trust your team. Learn how to delegate. Get a real nose for what's important, what's going to move the needle. I think that comes with experience. It's very difficult to, to, to teach that, you know, you know, be super focused on how your business makes money.

55:45 Yeah.

55:45 And what moves the needle with respect to that and what role you can play in making that better.

55:51 Yeah.

55:52 So kind of, you know, classic top down, bottom up views of the world I think is what, what you need to be a successful cto. Yeah.

56:05 Amazing. Right? Well look, we're coming towards the end, so I've got some quick fire questions to fire at you. First one is which brand or company do you most admire and why?

56:16 I'm not, I'm not a huge fanboy.

56:18 Yeah.

56:19 Organizations, I mean a company who I've, I've worked with for many years and we, you know, today I think I've done agreed up with databricks.

56:29 Yeah. Okay.

56:29 I'm a big databricks fan.

56:31 Yeah. What is it about them?

56:33 I think, I think the, you know, they have, they've grown hugely over the last few years. I mean I first interacted with those guys many, many years ago when I worked in the Bay Area, but they still kind of retained that, you know, they had the technologist choice. They solve the right problems, they take away things which slow down innovation and make them easy to do.

57:06 And they've done that through the world of cloud computing and big data and there's some cool stuff as well, applying those same methods to doing AI at scale, your real applied AI. So, so I think they're doing a great job. Yeah.

57:24 Amazing. What's the one piece of advice you wish someone had give you, given you when you were first starting out?

57:35 I mentioned this before. I think it's always try and see the world from other people's perspective that you deal with professionally, day to day.

57:47 Yeah. Do you think lots of people don't do that then when they first start out? I kind of get a bit too

57:54 much tunnel vision or perhaps they attach themselves to one viewer, one person's view rather than looking more widely. So I think, I think you've got to sample widely and then sample widely and then make your own opinion about how things work and what, what advice or what, what road you should take. Yeah. Yeah. Good.

58:20 If you could swap jobs with anyone for, for a day, who would it be and why?

58:28 I mean, I, I'm, I, I, I'm a wannabe guitarist.

58:32 Okay.

58:33 So I'd probably be some musician.

58:35 Yeah, yeah.

58:36 You know, maintain, but, but you know, the idea of, you know, being on a huge stage, playing guitar to an audience of tens of thousands of people. Yeah, definitely. No, that's, that's what I'd like to do.

58:46 Favorite band?

58:49 I, I think my favorite band are the Smiths. Yeah.

58:52 Yeah.

58:53 So I'm an old 80s and 90s indie guy, so. Yeah.

58:57 Nice. Are you kind of favorite business related book? You big reader?

59:02 Yeah, I'm not, not a huge reader of business related books.

59:05 There's a few non fiction, I guess.

59:07 Yeah. I mean I think my favorite book, and it's probably a bit specialized, is how to Measure Anything by Doug Hubbard.

59:16 Does that come back to the kind of thing you were talking about at the start around what kind of, you were, you were kind of fascinated.

59:21 Yeah, yeah. So he, he came out of a background of I think doing logistics for the U.S. marines.

59:27 Right.

59:28 And you know, had to optimize processes, make those things work. With very, very little data and you get a framework of how to quantitatively think about absolutely anything.

59:41 Yeah.

59:42 And I think it's a, you know, for. For, you know, working in. In, you know, insurance or anywhere else where you're trying to get quantitative. It's always good to have a back of a envelope view of what a number should be.

59:55 Yeah.

59:56 And. And, you know, Douglas Hubbard sets out to do that and that books. It's a great book.

60:00 Yeah. Nice. What do you think is the most important leadership trait or skill?

60:08 I think it's been able to pick the right problems to solve.

60:14 Yeah. Okay. Why?

60:20 Because it's very, very easy to. Let's take a step back from that. One of the things that I think most annoys me in businesses generally is what called premature optimization. It's throwing lots of resources inefficiently and doing a lot of unnecessary work that isn't really moving the needle.

60:52 Yeah.

60:53 I've seen that happen in many. That happens in all companies all the time. And there's a lot of theater around that there could be a lot of false metrics around showing success, but you're not solving the right problem and it's probably not moving your bottom line.

61:09 Yeah.

61:10 For the business. So I think in order to maximize your efficiency and also to maximize. People want to be doing something successful that makes a difference. That's a huge, you know, that's a huge motivator for.

61:23 Yeah, definitely.

61:23 Choosing those right problems to solve, I think overcomes us and it's. It's really difficult to do.

61:29 Yeah, yeah, yeah. I guess there's no exact science to it, but last second from. Last famous person that you most look up to or admire.

61:39 Again, I'm not, I'm not a huge.

61:42 Yeah.

61:42 Fanboy of people. I mean, I think, you know, there's been some. Some of the sort of unsung heroes that, you know, have sort of in the background built the world that we live in today.

61:57 Yeah.

61:58 There's a guy called Claude Shannon who worked in Bell labs in the U.S. you know, he really developed, you know, how to. How to transmit data, how to encode and decode signals, and it sits at the heart of everything that we do now. You know, some engineers and scientists probably heard of him, but beyond that. No, but, you know, he's one of the inventors of the modern world and no one knows about him.

62:25 Yeah, yeah, yeah. Maybe the best way to be.

62:27 Yeah.

62:29 And then the final question, as always, is what's the best thing about working in insurance?

62:36 It's. It's A very, you know, dynamic industry. I think I spoke earlier around. Some of the timescales in insurance are long, but some of them are actually quite short. And for certainly the role I am in envelop today in my team, we are actually able to see that what we do affects decisions straight away on a daily basis, which is very motivating and it's great because it's a great feedback to improve things.

63:11 I really enjoy working in that environment.

63:13 Amazing. Well, look, we've come to the end. Thank you so much for making some time to. To chat with us. It's, it's been great. I'm sure there'll be some people want to get in touch LinkedIn. Okay. If they, if they want to reach out and, and connect and. And yeah. You mean that is your offices in Bristol? Your office is in Bristol?

63:32 Yes, we've got office in Bristol and we've got offices in the city as well.

63:34 Well, if you're a budding engineer or something in Bristol, then. Then maybe get out. Absolutely, yeah, absolutely. I know they're hard to find a little plug for you there, but yeah, everyone like comment, subscribe, get in touch with myself and John if, if you need to and we will see you again next time.

63:51 Thank you.

63:52 Cheers, John. That's it for this episode of beyond the Desk. Thank you for listening and I hope you enjoyed it. If you did, please like the episode and share with someone in your network that you think would enjoy it too. It really does help us grow the show and reach more people. And remember to subscribe so that you don't miss any of the great guests that we've got coming up in this series. Beyond the Desk is sponsored by Invector Group, the insurance leadership search expert.

64:21 If you're looking to make a business critical or senior leadership hire and need support, we'd love to talk to you. Connect with me on LinkedIn or visit our website at invectorgroup.com for more information. I'll catch you next time on beyond the Desk.