%202.png)
This episode of Making Risk Flow is a recording of an event hosted by Cytora and InsTech to bring industry stakeholders, including clients and key industry executives, to the table to discuss how they are approaching digitisation. Juan is joined by James Platt, Henry Withinshaw, and Jonathon Gray to explore how digital transformation is reshaping commercial insurance, tackling challenges like reducing quote turnaround times, improving data accuracy, and enhancing broker service.
The panel shares real-world insights on balancing automation with human expertise, overcoming legacy system hurdles, and leveraging AI to streamline underwriting. Whether you're an insurer, broker, or tech enthusiast, this episode delivers actionable takeaways on modernising underwriting workflows for a more efficient and data-driven future.
Listen to the full episode: here
Juan de Castro: Hello, my name is Juan de Castro and you're listening to Making Risk Flow. Today's episode is a very special one. I wanted to share with you the conversation from a panel I was chairing at our quarterly live event a couple of weeks ago in London. We had three amazing panellists from whom you will hear later, and they brought a great mix of experiences and case studies from the transformation projects and a very insightful view on how to approach the adoption of new technology. These events always bring me so much energy and the sense that we are making massive steps to transform how we do commercial insurance. I hope you will enjoy this chat as much as our audience did. Welcome, everybody. Thank you again for coming. So we do these quarterly events, well, for two reasons. One is to bring like-minded people together, have a conversation with some of our clients, some of our key executives in the industry so that we can share lessons learned about how each of us is approaching digitization. And the second thing is also to show the latest features of our product. So I think that is how we're going to structure the session. We're going to do around 40-minute panel discussion, and then we're going to go straight into, half an hour demo. And hopefully this is going to be quite an engaging conversation, so please ask questions, challenges, comments. All of that is welcome. So with that, let's kick off the panel. So if you want to join me, so I've got Henry and James. I'll ask them to introduce themselves in just a second on stage. And then we've got Jonathon on VC. Unfortunately, he couldn't make it in person, but he will be with us on the VC today. Jonathon, can you hear us?
Jonathon Gray: Good afternoon, Juan. I can, hopefully hear me.
Juan de Castro: There we go. You should have seen yourself how you appeared on screen. It was quite magical. So yeah, let's start with what I didn't introduce myself. For those of you who don't know me, I'm Juan, and Chief Operating Officer at Cytora. James, do you want to introduce yourself?
James Platt: So I'm James Platt. My brief background is I was a consultant for a long time with BCG. I then moved to the dark side, which was the insurance industry, and spent about 10 years with Aon, primarily as COO, trying to turn Aon into an integrated global business. And then latterly, I led the digital and small client segment for Aon. And then finally, I left, mainly because it meant that I could join the Cytora Advisory Board.
Juan de Castro: There you go.
James Plat: And I've been out of Aon for a couple of years now. Again, back to consulting, actually, and working across the industry.
Juan de Castro: Thank you, James. Henry?
Henry Withinshaw: Good afternoon. I'm Henry Withinshaw. I am the Chief Operating Officer at Newline Group, which is part of the Odyssey Group, which is in turn owned by Fairfax. I started out as a reinsurance underwriter, and I spent most of my career in underwriting. I moved over to become a Chief Operating Officer when I was working in Dublin in around 2011, 2012. And I actually kept both those jobs going at the same time. So I was carrying on underwriting, as well as doing the new CEO role within that company. Today, I look after our IT and change side of things, as well as our underwriting operations. I actually look after HR as well and compliance. So I do a number of jobs. I'm one of those quite versatile Chief Operating Officer.
Juan de Castro: Thank you for joining us. Jonathon.
Jonathon Gray: Thanks, Juan. And good afternoon, everyone. I'd love to tell you I’m dialing in from somewhere glamorous like LA or Australia, but sadly not. I'm in my kitchen in Leeds. Last minute call up, it was childcare, meant I couldn't get to London. So apologies about that. But I'm Jonathon Gray. My current role is Head of Digitalization here at Allianz, and a role whereby my remit is looking at how we expand our digital footprint and how we make it easier for our brokers to trade more complex and larger risks online. But I'll be honest, I'm only about three weeks into that role. And prior to that, I was a business lead on our transformation division, looking at how we digitize the new business workflow within mid-market and utilizing tools such as Cytora, which hopefully I can share some insights today.
Juan de Castro: Thank you, Jonathon. And I think, for those of you who've been to some others of these gatherings, I love starting these discussions with like, what's the view from the client? And I think that James, I mean, you've left Aon, do we still think of you as someone bringing the broker experience. So from that perspective, when you're thinking about how insurers are digitizing from a broker perspective, what would you like them to focus on?
James Plat: Well, before I do, can I just gain a view of the room so that we had this conversation before about what's the bias? Who's working for underwriters or from the underwriting side of our value chain? And what about on the broking side? Not many. And what about overall? I'm not sure everyone raised their hand, so I'll assume there's maybe some reinsurance folk. Okay, so I can talk broadly from maybe the client side and the broker side, but for those that put your hand up for that, don't pull holes in my argument. It's interesting when you say what matters, and I think the core things that matter have not changed, and you could all reel them off exactly the same, right? You know what matters is you want cover. You actually want consistency. You want a decent price. I mean, they're the same things that have worked across our industry for many years. But, and I think we'll hopefully get into some of the but to this, there are other things that also really matter. And actually the makeup and I think the roles that the folk in this room have make a difference, which are increasingly making it easy matters a lot. Making it easy to work with you so that value chains are simpler, transactions are simpler, et cetera. Speed matters of them all. I mean, these things again, somewhat obvious, but it's remarkable how hard we make some of them in reality. And actually not just being easy to work with from a personal side, but obviously being easy to work with from a technology side. And I will get into some of the things on that. But the one thing I'd say about brokers, which I think does get lost and sometimes gets lost in London particularly, is what does a broker actually do? And that sounds like a really stupid question. But I found that what do we actually get paid for? And we have a conversation, right, about broker costs, et cetera. But what does a broker actually get paid for? And most people, particularly in the London market, naturally think broker gets paid for doing transactions because that's what you see. But actually, if you look at where brokers spend their money, they spend it with their clients day in, day out. Huge amounts of cost with clients trying to help clients understand their risk, trying to work out what to do with it before you start on the transaction. And that's actually, I think, really fundamental for this conversation. Because the more help you can give a broker in doing their everyday activity, not just doing the transaction, actually, that does make a difference. But last thing, two big issues in our industry, and one we've spent time talking about one of them, which is the fact that our transaction costs have not gone down. We have been in that conversation before. And then the second one is, which hopefully we'll come on to, service across for clients is abysmal. And we can maybe get into why and what that matters. But it does matter, and it is terrible.
Juan de Castro: As we were talking about speed of service and service, one quick anecdote. So I was in the U.S. last week seeing a U.S. insurer who's going to remain unnamed. Their quote turnaround time was 27 days, which it's quite amazing that you can still bind something if it takes you 27 days to produce a quote. I think the more interesting insight is they were facing significant challenges with the profitability of the book. And the hypothesis is the only risk they were able to bind 27 days late were the risks nobody else wanted. So I think sometimes we think about service as, I just want to make it easier for the broker, but it's also service actually ultimately correlates the quality of the risks that you can bind. That was quite an interesting insight. Okay, so with that context from James on what matters on the broker and the client side. Well, Jonathon, let's start with you. Let's start at the high level. When you're thinking about the digital transformation of Allianz, what are your objectives? What are you solving for? And where are you in that transformation?
Jonathon Gray: Yeah, that's a good question. I suppose the easy answer would be trying to solve James's problems. Yeah, that's fundamentally what we're trying to do. But I think it's no trade secret that Allianz are trying to transform globally. And we're very much on that journey here in the UK. And fundamentally, what we want to do is simplify and digitize our route to market. So making life simpler for our underwriters, making life simpler for our brokers. And like most commercial insurers, we want to grow, we want to grow in the right areas, we want to do so more profitably with less resources, you know, everyone's in the same boat. But if I look more at what I've been doing with my team, and might be worth just providing a bit of context for those who don't know Allianz, we've got three divisions within our commercial business. The smaller end, you've got digital trading, which where brokers self-serve, you know, they come to us via an extranet site, put in their risk details, and a traditionally smaller business, more simple, but traded online. And then the top end, you've kind of got large corp, which is a big bleacher company, stands multinational, quite complex case, needs quite a lot of underwriting to it. And then you've got mid-market, which is kind of everything that falls in between, which traditionally, again, is manually traded. We've got a number of underwriters out in our regions. And the bit that we've been on a journey on within our transformation division in the UK is within that mid-market segment, fundamentally trying to improve our speed to market and drive our data insights. And utilizing tools like Cytora, not going to shy away from that, but we've probably been in a world where that 27-day analogy you just gave. I could probably find risks within our business, like 10, 27 days, certainly before we've been on this journey. And, you know, we were in a world where a broker would send in a risk submission to us. We have various teams going away, re-keying data from one platform to another, going off searching Google, trying to find out as much as possible they could about the risk. Before then, back and forth between different teams to get it into the right system, the right place, before an underwriter could actually underwrite the risks meaningfully. And then I suppose, yeah, have a meaningful conversation with our brokers. Where we are now, a couple of years into that journey is you've got automatic risk ingestion we're extracting data points we're doing some automated enrichment calls doing a bit of routing doing a bit of prioritization for our underwriters to try and get risks fundamentally in a decision-ready format for our underwriters so that we take out the manual legwork um for our underwriters and then put them you know feed what used to be unstructured data in a structured manner into our downstream systems which allows our underwriters to get on the front foot and spend time trading with our brokers and our clients. Yeah, I sit here today, like I say, we're a couple of years into the journey, but we'd made from a proof of concept and now got a couple of lines of business flowing through some of our digital platforms.
Juan de Castro: Thank you, Jonathon. Henry, you're slightly earlier in the process. Perhaps it would be useful if you gave a bit of context about Newline and then explain how are you thinking about the transformation?
Henry Withinshaw: Yeah, thanks, Juan. We're in an early stage of developing what we need to do to make life better for our underwriters. We are mostly Lloyd's business. We have an insurance company here as well. It is mostly complex risk. It is the high value, high end, complex type of risk that we're doing. And for us, servicing our brokers is everything. Actually getting a quick quote turned around for them, getting documentation for them as required, being able to write globally is very important. For example, in our clinical trials business, you've got to have access to that paper around the world that our clients demand, and you've got to be able to make sure that you can give it to them within days. I mean, within a few small number of days. So our underwriters really need to know what is it that they can have, which will be helping them to provide that quicker than they currently do. We know the technology is out there. We have not been fast at getting there. But now that we know that what the problem is to solve for what our underwriters are looking for, we want to get that data to them. We want that document extraction. We want to use our data warehouse, the information that they're going to get on their price adequacy ratios, to know that we're writing profitable business, how it looks within the overall portfolio, and improve how they're actually underwriting, as well as getting out those documents to their clients. All of those are absolutely essential to us. So we do it at the moment. We have workarounds for making this work. We're extremely profitable and have been for a very long time. But we know we can do it better. And so now is the time to do that.
Juan de Castro: Okay. So it's interesting, like, obviously, two very different organizations, but I've captured probably three themes that are at the core of both. One is how do you remove all those manual activities from the underwriters that are just consuming capacity? Second is how do you capture data to provide the best insights to the underwriters? And by doing that, how do you optimize your service and your speed to getting back to the broker? Probably those are three themes. That is quite interesting how, a very, very large organization and you guys are probably focused on the same point. And we've been talking about service. So I guess going back to you, James, from a broker perspective, again, service, we often talk about things like quotes during our own time. But how would you define service in a broader sense?
James Plat: It's funny, Juan that, I think I got, in my prior roles, I got asked that question, you know, how does a client, and we, you just say brokers are the proxy for client, but fundamentally, we mean end client, right? It's what we're really trying to get at. What does the end client care about? And oddly enough, quote, turnaround, and maybe let's talk large and sort of small and split them for a moment because they are so fundamentally different, right? I mean, obviously, both your businesses, different parts, different places. The large end, quote, turnaround matters, but it's probably, ironically, not always for large risk managers, at least in my experience, the question that's really bothering them because it's going to be complicated. That place is complicated anyway at the large end. There’s going to be back and forth. They want to get it faster, but it's not quite the same as a small client wanting better than 27 days. I mean, another just because stats matter. I did a survey a couple of years ago, and this may now all be different. And this is U.S., but policy issuance, which does matter, right? Because you could argue with the binder versus the policy, but policy certainty is pretty important for most clients. The average policy issuance was over 100 days. Now, the good news was when we looked into this and tried to work it out, we reckon 50 days were due to the underwriters and 50 days were due to the brokers. So at least it was fair. But policy assurance, again, is only part of the journey. Again, if you're thinking of the role of a risk manager in a large organization, you not only need to have your quote and understand it, make sure you've got the capacity, et cetera. You not only need to be confident that you've got certainty around your policy. You then need to have moved the money and have the invoices. And the real trick in this process is that you've paid the invoices before the next policy starts. And that sounds, that's only 365 days. And we had a number of clients with large programs where we were getting pretty close to 365 days in that process. So the one thing I guess, again, if you go back to the client, almost put the broker to one side because brokers are pretty simple. What they want is for their client to be happy, and for them to put a minimum amount of effort in. If they can achieve those two things, then they're pretty happy. But it's that end client and they need all these other pieces because they'll have their risk committee asking them these questions. That, to me, is what we as an industry are trying to sort of focus on such that broader end. Obviously, at the small end, and sorry, the other side of it, it's quite different. But it's similar. But suddenly, quote, response does matter. And I've just started this small business, tiny little business. And if you want an ironic story, so we decided that we're going to consult one of the large brokers. And you want to try and deal as a small company with the insurance requirements that a large broker creates. But they wanted us to get a huge amount of cyber insurance. Which, literally took two weeks for a firm that had three people to try and get the quote back. I mean, that is just insane. I mean, totally insane. So this is a problem at the small end and it's a problem at the large end.
Juan de Castro: Going back to you, Jonathon, I guess, as you said, you've been going through this transformation for a few years now. Can you give us a deep down, next level of detail of what the underwriting workflow looks like? And also, I think what would be quite interesting for the audience is lessons learned, like things that you got right, things that you got wrong. Any lessons learned?
Jonathon Gray: Yeah, of course. And I think for me, any target workflow, and it's probably an obvious cliche point, but for me, it's having the underwriter in that journey and in that workflow where they add value. For me, the use of technology and everything we're doing on transformation isn't around replacing underwriters. It's about working hand in hand with the technology. And if I look at where we are today, we've got tools like yourself in place, extracting data, enriching data, routing prioritization. Where I'd like us to get to is actually take that a step further, especially around that triage and being transparent. And depending on some of our lines of business, we've still got humans in the loop checking data points because there's still new technology. The world of AI, LLMs, et cetera, is moving. With certain points in our journey, we've still got humans checking certain data points. For me, that's a value destroyer. And if I look at that target workflow, do I see a world where I'd want to get to is having a system that extracts the data, enriches it, triages it, routes it to the right place. So, for example, if we've got a small property owner's risk, you know, one location, but in a decent area, not going to flood, claims free, relatively simple. That should be routed down to our digital workflow and it should go through no problem. On the flip side, if it's more complex risk. Where property owner of 10, 15 locations, some of those locations are fast food restaurants, making up. Now, you need a bit more of that underwriter touch and that's where the underwriter comes in and adds their volume. So, for me, it's having that underwriter look at the right cases and have to step in where they need to and having the technology remove the rekey and removing the risks they don't need to look at. I suppose if I think about lessons learned and, you know, the journey we've been on, again, it might be cliche, but I think it's start small and get it right. I think it's fair to say we had some success stories, had some lessons learned. We tried to tackle too much all at once. You know, is this get some technology provider in, let's look at the journey today, it's taking this amount of time, we want to get to X. We try to run and move really, really quickly. And I think if you look at the end-to-end new business journey, we change a lot of it all at once. And actually, yeah, okay, the overall metrics are good. But actually, I think we could have had even more success if we'd honed in on a particular part of that journey and got it right and got our underwriters on board and then expanded from there. So for me, it's I think, if I look at our technology, a transformation journey at the minute, I think all the building blocks are there. For me, it's leveraging and going further. But also, yeah, lessons learned, start small and then expand.
Juan de Castro: And to just build on that, so you said some data points, for example, are still not accurate enough. So what has been more difficult than you expected?
Jonathon Gray: Yeah, it's a good question. I think the pace of change is one. But I think if I was going to really drill down, I think it's people's adoption of technology has been a bit of a surprise. And it's a bit of a surprise to me. I mean we’d a meal deal at lunchtime with a self-service checkout in Sainsbury's and lost our cool with it because it just wouldn't work. But we see the same in underwriting. And if I think about how our underwriters are adopting the technology, I often use the analogy that the technology we've got in place is like a junior underwriter at the minute. It needs training. It needs experience. It needs a bit of hand-holding. If I was sitting in an office today with one of our junior underwriters and they looked at a risk and thought, this is the answer, I think it's certain trade type, et cetera, and I disagreed with that, I probably wouldn't turn around and say, well, you're right. And I probably wouldn't turn around and say, I'm never working with you again. But actually, that's kind of the feeling sometimes we've got in our business around technology when it goes wrong, we hear about it or the negativity. And actually, it's like, oh, well, the system's got it wrong. I disagree with that. I don't trust. And if I look at do, there are lots of examples, but two examples really come to mind around the use of technology and data. One, that's gone well and it's been really adopted. And one, we've got some challenges. So I'll just spend a minute or so just going through that to try and bring it to life. One of the things we're doing is we do an enrichment call. We look at what the occupancy type is. It's a fairly basic one-to-one lookup. And actually, we had a risk. I think it came through Southampton, for example. Broker said it's an office risk. Nothing to worry about. It's an office risk. Went through the system and actually the system identified as a nightclub. Now, I don't know if there are many underwriters in the room, I can't quite see you all, but that's a very different risk and actually led us to a different result in what we did in the journey. And actually, quite quickly, it went around our business. The technologies helped us identify and challenge our brokers. Quite a simple use case worked well. With technology advancing in the use of LLNs and Generative AI, we've also got a different use case where we're trying to use technology to understand what risk type it is and what our appetite is. And is this a good risk for Allianz? What it does is it, you know, the technology scans a document and tries and maps it to an industry code. And actually, I was called a couple of weeks ago and said, the system's rubbish. It's not getting it right. I looked at it and I think the presentation basically said the client makes and sells pasties, I think it was. And the system had gone through and it classified it as manufacturer of chilled foods or goods or something like that. That makes sense for the life of me. I was like, this all make sense, underwriter chosen, processing of meat o something along those lines and I was like, for the life of me, I just do not understand how the underwriter has chosen this. What happened? She, the undrerwriter picked up the phone to the broker. Had started talking about the risk, the googled it, looked at their website and actually saw that they made pork pies actually our underwriter determined that the handling of meat was a greater risk in the selling of pasties. And I think what that brought in life to me is a technology did exactly what we wanted to do. But actually, the expectation of an underwriter was different to what was happening in reality. And the answer back I gave to the underwriter was, well, this is where you're adding value. The system can't pick up the phone to the broker and have a meaningful conversation about the risks. But that's been a bit of a culture shock within our business and actually trying to bring underwriters on that journey around. This is what technology can do. This is where it's at. We need to feed it. We've got to learn it. And we've got to expand from there. So just two interesting use cases that jumped out at me.
Juan de Castro: I think actually those two resonate really well. And we see that across every single client. And I think these are the two sides of the coin. I think one is set expectations that data is not going to show exactly what you would expect for every risk. And there's a process of training. There's a process of providing feedback to any type of technology to drive accuracy up over time. I think the other thing that really resonated is you told the story about the nightclub. I think we can all do more of bringing those kind of exciting stories and making more noise about those. Because sometimes that becomes just an anecdote that a few underwriters know. And I think you have to do, I think, the balance of the two. One is set expectations that it will improve over time. And the second one is when there's a big win driven by technology, also publicize it and make more noise about it. But Henry, do these two things resonate with you?
Henry Withinshaw: Yeah, very much so. And it's exactly the same that Jonathon spoke to do is bringing those underwriters into those conversations from a very early stage so that they are involved and remain involved from start to finish and going slowly at it. We developed last year an underwriting data extraction tool that was going to help us with our U.S. domiciled exposure. Our reinsurers were concerned that there was probably more life science, particularly and D&O business in the States, which was domestic. And we were saying to them, no, it's not. You know, our U.S. D&O is really ADR, it's a jointly listed business. But we'd had a couple of claims that came out of the U.S.. And they were basically going to charge us in our reinsurance program for their guess what our us exposure looked like. So we built a tool. We built a tool with underwriters. We got it to them quickly. They were involved in the process the entire way through. And that tool probably saved us about two million pounds in our reinsurance this year. So if you get it right and you get the buy in, we can all see the rewards from it. But you've got to get it right and do it right. I think most of it also is about being pragmatic with these things. Don't try and boil the ocean. Don't try and solve all the problems that everyone has. We can get that right in the second stage or the third stage. And we're now going through that process of reinventing that tool ourselves this time round. But as we go at this sort of augmentation of underwriting, data extraction, data from our warehouse, policy issuance, and all those pieces, claims into that process as well. We won't go at that and rush a bit, and try and solve all the problems at the same time. We'll learn what we did with our U.S. exposure stuff. And we'll do it in the same way. Because actually, what the underwriters really care about is they've got a proven technology that works for them. And so often, I don't know whether you said it, James or Jonathon, so often our underwriters feel let down by not having delivered what they asked for, et cetera. And they give up on us. So it's good to have some good news stories.
Juan de Castro: Fantastic. Yeah. I want to make sure that we've got a few minutes for Q & A. But before doing that, we've been a bit hard on underwriters. And how inefficient those processes are, et cetera. But why not do the same with the broking side? You said on brokers, I mean, 50% of those 100 days were on the broker side. So how are brokers thinking about how do they transform? How do they operate?
James Plat: Well, it's interesting, because I think, I guess I'm a believer in the issue normally is always upstream. So a lot of our issues, and you want to blame the underwriters, et cetera. But the issue is, do the underwriters have the data from the brokers at the right time with the right accuracy? And ironically, you can then go one step further. And say, do the clients provide the data to the brokers with the right accuracy? And if you look at the worst lines often for service and performance, it's things like real estate, where the data is changing, obviously, rapidly. And it's very, very hard to keep up with it. And that creates issues the whole way down the value chain. So for me, if you're a broker thinking about this for a moment, it's pretty much the same challenges, right? If you are creating the submission to provide, you've still got to bring data in. You've still got... It's the same thing. And ironically, and this is what kills us, industry-wide, it's the same data. Fundamentally, you're going to add data to it, right? I mean, you're going to need to put more in for underwriting purposes. But the core data is the same. It annoys me that we haven't been able to create some form of structure that everybody uses that really is common. And we just pass that data across. And we struggle, well, brokers struggle. I struggle now as a broker, getting the data out of the client, often wrong, often in terrible formats. And then we struggle to get it into a decent place to pass it on to the underwriter. And when stuff comes back from underwriters, it's normally not what the broker asks for exactly, right? So the underwriter, well, I didn't quite want to do that. You know, I changed this a little bit. I hope that was okay. And then the broker's got to unpick it and provide the client. So this is a systemic issue. I think brokers are putting a huge amount of effort into their submission placement processes, the pre-buying stuff that they're really trying to get right, hopefully partnering. And if they can get that right, and we can agree some sort of sensible way, and I think it does unlock for the underwriters to be much more efficient. But the thing of all of that that has really given me hope, and you talk about LLMs, I honestly think the biggest use of LLMs that we could put would be, I hate this, everybody hates digital technology, right? Raise your hand if you actually like going onto a platform and typing in stuff and using it. Nobody does. And that to me is really, really fascinating that everybody hates digital stuff, yet we spend our whole time trying to, to implement technology. But LLMs give us a chance to be able to literally do it. Imagine the meeting that you're having between the underwriter and the broker, the LLM or the generative AI, whatever, is capturing that information and putting it straight into everybody's structured documents with no filling in. When you can do that, Cytora, I think we'll all be.
Juan de Castro: That would be the next quarter. I think you're totally right. It's like quite a common conversation with clients is if brokers only sent us all the information at once.
Henry Withinshaw: Yeah, even the master form contract version three, which was supposed to be a standard contract, standard format in a digital version across the London market, that never got off the ground. And had that come out across the market, we would all be able to ingest it nice and easily into our systems, but it hasn't got going. Now it will do, and it should do, and Blueprint 2 should help these things happen when it gets going. But we do need that push towards that standardization.
Juan de Castro: I fully agree. Okay, we've got a few more minutes. Any questions or comments? I mean, I know most of you guys, and all of you are facing similar questions, challenges, thinking about how you're thinking about your digital transformation. I think this is an opportunity to ask questions to the panel.
James Plat: Or the panel will ask questions of you.
Question 1: James, I was interested in your point about real estate property data being the worst. Henry Withinshaw, you've got a business that looks at liability risk. Jonathon, I think Allianz is mostly property. It's interesting if we can put a positive spin on this. Where do you see the lessons from the different data types as to where we can get standard data? Because I think, James, if you're saying property is a hard one, then we're really stuffed, because that should be the easiest one, because property generally doesn't move around. You can define it. So I'm just interested in what the panel's view is on these different data types and some positive outcomes.
James Plat: And just on that, just to be clear what I was getting, it was our real estate clients, where they had constantly changing property data, which was the biggest killer for us. And I literally ended up having weekly phone calls with our New York office, because all the processing was screwed up. And I got bored of being beaten up at the time, and I got beaten up a lot. But it turned out the clients never really provided the correct data. And so eventually, we found mechanisms to do it. And I feel that, I mean, there were a couple of technologies out there, right? Sort of, and a big ticket of trying, to do things in this sort of space to improve that. Pelago, if I pronounce it right, I'd be biased to either. And I'm sure others get in that data. But property, you're right, should be easy. But I don't know other challenging types.
Jonathon Gray: Yeah, it should be easy because you should see it. And I think from an insurance point of view, I think we recognize, yeah, property owners or real estate, we do have similar struggles. And just in terms of how brokers present the data to us, I mean, what one piece of data that really jumps out to me that we'd love to get our hands on is basement data. Fundamentally, geospatial property, you can see it, but basements, you can't see. And actually, if you think about the flooding risk across the UK, you become quite exposed quite quickly, just as you can't see. It feels like as technology moves, there should be databases available around this kind of stuff. And we've seen certain players and certain insurtech providers looking to pull this data together. But yeah, we have the similar struggles. And I think it's trying to get your hands on as much data as possible. The right data, of course, being able to trust it and on the right, be geocoded correctly to the right location, because even that throws up different anomalies. But yeah. So base on data is one that really jumps out as a particular use case.
Juan de Castro: Good. Any other question over there?
Question 2: More for the underwriters, if that's all right. If the engine room of any underwriting operation is your policy admin system, I'm just intrigued to know how old your policy admin systems that you're using are and how much that's stopping you from progressing the innovation within your organizations.
Henry Withinshaw: I'm happy to take that one. So our policy administration system, we have no legacy systems. We used one system from the start when Newline started underwriting. We upgraded it in 2015. So we had this version two, which is what is still being used today. And last year we did a major upgrade on that system to basically get it ready for Blueprint 2 and digital risks and API connectivity, eBot, eCot acceptance, et cetera. And we are very lucky not to have legacy. Everybody complains about it, but everyone complains about most PaaS systems. So I'm not really surprised. What we have to do is get to a point where the PaaS is in the background and not in the foreground. The underwriting, the pricing tools, the getting the claims information through the document extraction, the price adequacies and all those things, the exposure changes year to year. That needs to sit at the front of what the underwriters are looking at, not the policy administration system. So that's where we need to get to. And an integrated one that does have API connectivity is obviously going to be very important. We won't be changing it though because, I really don't want a legacy system.
Jonathon Gray: I mean, I wish we were as lucky as Henry, and not have any legacy systems. We still got some green screens floating around our business. I remember joining 15 years ago when they were there. They're still here now. But that is a transformation we are under, is underway within Allianz. It's end-to-end across our rating tools, our mainframe system, our claim system, etc. So, unfortunately, yeah, we're very much in the same, you know, we're trying to do exactly the same as Henry articulates, have things all integrated and connected in a seamless way. But it certainly throws out its challenges. Which is then impeding speed back to the broken point.
Henry Withinshaw: Yeah. It does if you are relying on your PAS to do it, but we won't be. So the PAS will just be holding the data, the record, and hopefully ingesting it automatically as well, digitally, without having people, again, having to re-key it, and providing the data into the general ledger to provide our financial reporting, but not doing much more than that. Very happy for the pricing to be kept outside of the system. And really, it's the data warehouse is, again, extracted out of the policy administration system that is going to be the live tool for how we're getting on with our business, how it compares to last year, where the profitable lines are, where the opportunities are, where we need to grow, where we need to pull back. So the PAS is literally, it's going to be far more at the back and less at the front as it has been in our lives in underwriting.
James Plat: I think what's interesting at this point around, it's funny, it's what the banks went through many years ago, right, when they were faced by an online world, and they had these horrific banking systems, they had to sort of deal with it, and how did you get them to the back? And I think they went through it earlier. But again, what I think has changed, and you mentioned the API, as long as a system can take some form of API, new data interconnect, data technology, and the ability to move between front and back has completely revolutionized it. And we were going through a process of literally looking at every system. If it had the ability to API keeping it, and if it didn't, we'd have to change it ultimately, and then building new on the front. And I'm pretty sure that was the broking side more than the underwriting.
Juan de Castro: James, Henry, Jonathon, thank you so much for joining today. Hopefully, it's been interesting for everybody. Thank you. Making Risk Flow is brought to you by Cytora. If you enjoy this podcast, consider subscribing to Making Risk Flow in Apple Podcasts, Spotify, or wherever you get your podcasts so you never miss an episode. To find out more about Cytora, visit cytora.com. Thanks for joining me. See you next time.
