Insig AI (LSE:INSG) — Investor Presentation

0:01 Good afternoon, ladies and gentlemen. Welcome to the Catena Group PLC investor presentation. Throughout this presentation, investors will be in listen only mode. Questions are encouraging to be submitted any time via the Q and A tab situated in the right hand corner of your screen. Simply type in your question and press send. The company may not be in a position to answer every question it receives during the meeting itself. However, the company review all questions submitted today and publish responses where it's appropriate to do so. These will be available via your investor me company dashboard, and we'll send you a notification once they're ready for your review. I'd also like to remind you that

0:33 this presentation is being recorded. Before we begin, we'd like to submit the following poll. I'd now like to hand you over to Matthew Farnam Schneider, CEO of Catena Group and Stephen Cretna, CEO of INSIG AI PLC. Good afternoon. Hello, and thank you, Mark. Hello to all of you. Thank you so much for taking the time to to meet with us today. As as Marcus said, my name is Matthew Farnam Schneider. I'm the CEO of Catena Group. And our purpose for wanting to meet with you today is to give you a bit of background on the transaction that we announced last week and our our acquisition of INSIG AI PLC. As Mark has mentioned, we we have with us Steve Cracknell, who's the CEO of INSIG AI PLC.

1:20 We got to know INSIG AI PLC, or I I got to know INSIG about two years ago, and I've been incredibly impressed by the work that Steve and his cofounder, Warren Pearson, have have done in setting up this company and the incredible work that they've done in helping asset managers to make the transition into a data led, a data centric investing strategy. And I think that's that's something that you're you're going to hear a fair amount about today from Steve, and it really is a a a quite critical step change within the asset management industry and and and one that we are, realizing and and and many, asset managers are are realizing is a critical component to their business going forward.

2:12 So, really, what interests us in, in INSIG Capital, more than a year ago when we made our our initial investment in them, was the the the combination of, of data science,

2:27 engineering, and and and and and very clear sectoral knowledge that the INSIG Capital team have brought and the success that they have had with helping asset manager clients to make this transition. So that was the rationale for Katina Group to make its initial investment in INSIG just over a year ago. And now the rationale for the acquisition of INSIG Capital and and and very much our focus on developing this business and and helping helping the business to help its its customers and future customers to make this step change, to make this transition.

3:12 So with that all said, I will turn you over to Steve Cracknell, who can speak to you about the the the history of INSIG Capital and some of his background as well as a little bit further about where we are taking the company going forward. Thank you. Great. Well, thank you, Matt. Well, I'd love to say it's good to meet all of you, but we can't see anyone. It's good to be sitting here today. Yes, just to, I guess, contextualize why INSIG Capital started, it goes back just over three years ago. It was a collection of, I guess, skill sets that myself and Warren had acquired over almost two to twenty years of working at Goldman's in financial technology, building financial systems, building trading systems and learning, I guess, a rigorous way of building hardcore software that's repeatable and scalable.

4:09 The second part of that, the story, which has ended up running a machine learning company in Silicon Valley and moving out of a big corporate banking environment into the West Coast Of America, you're exposed to cutting edge technology, cutting edge cloud, innovative machine learning technologies. And what was I guess interesting through the experience is that if you combine the two, the one bit of discipline around what I would call the banking expertise and then the Silicon Valley creativity, you can create a very unique combination of product offerings. And this is something that we assessed to be lacking in the asset management industry, an industry that's typically made a lot of money.

5:00 They've got a group of very smart investment professionals but haven't had the opportunity to take advantage of modern tech. And in 2017, we said this process of being able to take advantage of a cloud environment requires something called DevOps or development operations. Or another way of thinking about it, it's plumbing. How do I get data from my computer onto a server really quickly? And then if I want to do a computationally heavy algorithm, how do I scale from one computer to a thousand computers? And that plumbing piece is an important component. The next segment to that is to say, well, what's the business driver? Having a bunch of data scientists and engineers in a room together might generate a lot of interesting results, but are you going to make money off the back of it? So you need that business driver.

5:56 When you combine the business with data scientists, with the engineers and the cloud component, you get a very unique offering and that's kind of where INSIG started. Our starting point was to develop suite of machine learning classifiers that would read through earnings call transcripts and pull out what direction companies are going. Are they focusing on technology strategy? Are they focusing on cost cutting, m and a? And we thought we were really clever. And we went and pitched this to a bunch of asset managers. And they said, wow. This machine learning and AI stuff is phenomenal. We're not that interested in your classifiers, but could you do that for us? Could you come into our organization, help us build this cloud based infrastructure and transform our data into something that's useful and actionable to us?

6:51 And that's really where we started. What we learned very quickly was that the skill set that we brought was in very high demand. We were able to charge very good consulting fees and within twelve to fourteen months, we were a profitable business. And we grew quite quickly in the first eighteen months and simply doing the same thing, build an infrastructure that can support scale, integrate machine learning with a business overlay. And that was the kind of, I guess, offering we were providing at that point. Through the consulting experience and developing solutions for individual clients, a theme started to recur.

7:38 How do I get third party data into a machine readable format? How do I integrate custom or proprietary data with third party data? So we built a data framework. How do we get data out of Excel into Python? How do we get data out of Excel and convert the Excel formulas into Python code into a database? So we built Excelathon. Once you've done the data transformation and once you have that information in a in a in a database, how do you optimize it? And then how do you then overlay machine learning algorithms to look for alpha opportunities? And this is where the INSIG portfolio or portfolio insights is a product that started to develop.

8:23 Interestingly, when when when we first met Matt, he said to us, hey. You guys are doing really well as a consulting business, and you are effectively providing products to your clients.

8:36 Why don't you guys become a product company?

8:38 And, initially, we played a bit hard to get, and we said, well, this consulting gig is doing okay. We're generating good revenues. But I think, to be fair to Matt and kind of his vision on where the growth potential was, they said, look, as a product based company and a company that know already knows how to scale based on our our kind of, you know, historical experience, the scope for growth is massive. And, you know, if if if we can help you take your skill set, productize it,

9:11 potentially, we can grow this into a big company. So that's where the conversation started, and that's also where the initial capital injection came just over a year ago. And I guess, coincidentally, we'd, I guess, verbally agreed to start focusing on product and and and taper off from consulting. And I and I say coincidentally because that's also when when when COVID kicked in. And it's very difficult to sell consulting services when you can't go and visit or see any clients. So we really did focus our all of our efforts, all of our engineering skills on building each of these products over the past over the past year. Fast forward to the end of twenty twenty, the demand we were seeing from asset managers we're talking to was immense.

10:02 And the same questions were coming up, you know, how do we get our data in? How quickly can we get signed up? What are the optimization strategies, you know, and and that's really what solidified the the decision for us to to merge and join forces and and and list INSIG AI as a new machine learning business. I mean, I'll pause there for a second so I can have a sip of water. Don't know if you have anything to add to that. No. I think that's right. What I put on the screen here is an outline of the the INSIG products. As Steve mentioned, the the core the the core products at this point are the INSIG portfolio and INSIG data products. Those are the ones that that are in production and and being used by by clients.

10:53 One of one of the products that we're we're most excited about during the course of the coming months is the INSIG ESG product. And Steve can speak to that better than I can, but but effectively, it's it solves it it it helps to solve what has been a quite critical and important challenge within the asset management space. And that is how do we how do we define an ESG compliant universe of of of companies,

11:27 and how do we do that credibly? A number of other entities have have attempted to to provide ESG scores and and some sort of confidence around ESG compliant companies. But much of that is being done in a very opaque manner, and you're relying upon you are are essentially leasing the confidence of another provider, MSCI or someone like that.

11:57 Whereas the way in which Steve and the team have have gone about addressing this challenge is a data centric method. I I really looking for what evidence can we can we provide that a particular company is either ESG compliant or not. But then also, how can we do that in a manner that enables any individual investor, any individual client to to apply their own structure, their own sort of waiting system to what they perceive to be ESG compliance. Because we we've we've recognized, and I think much of the industry is recognizing that ESG can mean something different to different investors.

12:44 And and their appetite for certain types of environment, social, or governance risk will be very different according to whichever individual asset manager or even investor you you speak

12:57 to. Yeah. I mean, that's I mean, that's I guess that's a very good description, Matt, of the the thought process and use the phrase data centric. You know, I guess an interesting anecdote around machine learning and AI is that it's it sounds very interesting. It sounds very exciting. The promise of it is is the appeal and and and what people want to talk about. But when you step away from it for a second, in order to get the value out of machine learning algorithms and artificial intelligence capabilities, you have to have the data.

13:34 And the data has to be cleaned, it has to be transformed, and it has to be stored in a machine readable format. And that process is incredibly difficult. It's not pretty, but it has to be done. And, you know, in each of the products that we've developed, the theme is exactly the same, is you have to do the heavy lift around the data transformation first. At the point in which you have that information stored in an accessible database, that's when you can do the interesting stuff. In context of ESG, the starting point was let's go and see what's out there in the market today. As Matt said, you know, there's lots of scoring systems. There's some very incredibly reputable providers of scores and they've done their own analysis. But they have to get the same underlying data, same underlying metrics, the same annual reports, earnings call transcripts, sustainability reports.

14:26 Now from an asset management perspective and regulatory requirements, they have to stand behind their investment decisions and they cannot just rely on an independent score from a third party provider. So they have to validate the information. And as such, have to then go and read each of those documents and each of those reports. And if their investment universe is a thousand companies, you now need to read a thousand annual reports, a thousand sustainability reports.

14:54 So our starting point was to say, this information is already in our data base and we can mine it. Let's build a suite of text based machine learning classifiers to be able to surface evidence of ES and G issues. And if we think of environment as a pillar, the issues that fall within the environment are climate, GHG emissions, waste and the governance that's around management and control. So we've created a suite of 15 discrete machine learning classifiers that can speed read these documents. Now the value of that is, I think, relatively straightforward.

15:40 I have a universe of a thousand companies. I have three sets of documents and I have 3,000 documents to get through. The machine learning classifiers will surface the evidence that is relevant and necessary to validate what a company is doing or is not doing

15:56 as the case may be. The second part of that is to say, well, if the classifier can get me to the content quickly, I then want to interrogate it. So that's where we do data engineering and we do smart software development where you can actually click through and you can get down to the actual core or underlying data document. This idea of the talk, that's I guess what the classification part does. It's let's identify what companies are talking about. The second part, which Matt mentioned around the metrics, it's the underlying metrics that are published, what are the targets, what are the goals that companies are trying to get to, what are their emissions, what is their usage of water.

16:44 And that database part of it or the numeric part of it is getting tied together with the the the text based part. And what you end up having is a research tool that allows you to see what a company has said and validate whether it's actually what they've done. And the feedback we've been getting from from our asset management, I guess, inner circle of of good clients or close clients is that it's exactly what they need at this point in what is quite a difficult ESG environment because the thing you know, the requirements are changing constantly. The standards are changing constantly. And this second layer of interrogation has a high demand at the moment.

17:42 I've I've just turned the slides back to the ESG product summary slide so that so that you can see a bit about how the the product looks. And and and to note, this is a product that is is

18:00 is in a a very early alpha stage at this point. So we're we're we're having conversations with a number of of existing consulting clients and and a number of other interested parties on on the early stage of this product. But we we look forward to launching it fully by the third quarter of this year. Yeah. Yeah. I mean, actually, the another, I guess, belief we have as a business is that get it. Someone's trying to call ask a question. It's around how we develop the software and how the software should look. In what we call a modern tech world where as consumers of things like Uber or Deliveroo or Amazon, the applications we get used to using and being able to log into on our phone, the usability has become so seamless and easy.

18:58 Part of what we believe in is that the tools we develop have to have that same kind of usability and that the days of antiquated software and very difficult to use systems isn't going to cut it. That is also part of the challenge to do something and make it seem easy to do is engineering from an engineering perspective quite is very difficult. But we think it's absolutely critical to the success of our business Because if we can't make machine learning models accessible and we can show our workings and we can show people what the models are doing and the assumptions that they're making, we're not gonna get people's confidence to use those systems. So that's that's part of what our delivery is, is that we show what we do. And the tools that they can access it through, which is it's it's a web application, has to look nice and it has to feel nice and it has to navigate

19:54 well. And we think that's really the new standard that certainly financial tools will have to meet. We hope to at least set the bar for what will become accepted and expected by investment professionals. But, again, it's it's having this combination of the hardcore engineering and the data science to to interrogate and and surface the important data. And then the second part of it is to make it easy to read and interact with. So, again, that's, I guess, part of our our product mandate is to make sure we we deliver a a product that's you know, gives me the information I need to make an important or not important, but a good investment decision and looks nice as well. I think that's that's really where we're balancing ourselves. Well, I think something that Steve just said there is is quite important to reinforce, is the the INSIG AI tools are not intended to supplant asset managers or or or fund managers' investment decisions themselves.

21:04 They're meant that they the the tools provide further information to augment the strategies, to help them to help them to to gain value from vast amounts of data and and and data that in its current form often just becomes noise. So this provides a structure and an ability via the the machine learning tools, via the the front end applications to to gain value from that. And and I'm just gonna turn over to a slide that outlines the the INSIG portfolio product, which perhaps Steve can can provide a little bit more detail around exactly how this product works and also the the development path that that we see for it over the course of of the coming months.

21:55 So yeah. So the I guess the some of the feedback we had early on around the the application or usage of machine learning models is that asset managers, whilst they think it's really interesting and exciting to do, they don't want an answer that comes out of a black box. They don't want to pin their investment decision on something that they don't understand and that they can't control. And, you know, in in the context of the the visual image of a black box, if you give all of your information to a data scientist, in this it's the same thing. Now you're entrusting another person to tell you what the right thing to do is. And very early on we decided that when we run machine learning models over data, we will show the assumptions that those models made, we will show the weighting, the scoring, each of the components that was used. And the minute we, I guess, surfaced that information, the trust that it engendered through those investment professionals was was immense because for the first time, I could say, I understand what this is doing.

23:00 And then the thing that we saw, which I think really is going to differentiate our product, is that they have the opportunity to overlay their investment expertise and their flavor and their experience of what they know actually is going on in the marketplace. And and, you know, so for example, in this this image, it's showing, you know, three lines which have outperformed the benchmark, and those are three independent machine learning models. Now we're not prescribing which is the best model to use. We're simply saying these models have made assumptions based on twenty years' worth of historical financial data, which by selecting a portfolio of companies has outperformed the benchmark.

23:40 That's step one. Step two is to say, okay, let's go and see which assets those are, what is their distribution, you know, what sectors do they fall within, you know, which, you know, which regions do they do they make up, And again, showing all of the underlying workings or the data that underpins it allows them to then say, okay, I'm now going to make a more confident investment decision.

24:02 The other value of the application of the models is that, obviously, we can't look into the future and we can't have perfect foresight. But when the market changes, because the models can run historically over similar market periods, it helps, you know, investors understand to say, okay. When we were in a high volatility period, what did it look like then? Which companies were selected then and why? And then that helps inform their next decision or the next part of their portfolio construction process. And I think it actually, that also is a very important point that what we're seeing from quite a few of the existing clients and and a number of the the potential clients that we've been speaking to is that they're starting to recognize the speed at which the markets are adapting.

24:51 And their their prior their prior operational, structure for adapting investment strategies was potentially too slow to, to match the pace of the adaptation of the market. And so these tools, the the INSIG portfolio tool in particular, gives them, gives them a a a suite of tools and insights that allow them to more rapidly adapt to the the changes within the market and to to recognize those changes more quickly than they they could have previously. What I might do is just turn to

25:31 the other slides on on on the three other products that perhaps Steve can just walk through very quickly. And then I I know that a couple of questions are coming through, but perhaps we will work through these and then turn to some questions. So, I mean, we mentioned the data product, which is really a data transformation framework. It takes data from third party sources and then this notion of cleaning, transforming, making machine readable. It's something that's fundamental to any first step to being able to do machine learning or or AI. Offered as a framework, it means that, you know, asset managers that haven't traditionally had technical teams or engineering teams, they then don't need to go and hire a data engineer, DevOps engineer, and a data scientist to get over that first hurdle.

26:26 So this this framework really facilitates them, you know, saving probably twelve to eighteen months worth of effort to get the benefit of some kind of machine learning output. On the on the the docs side of of the equation, again, coming back to this challenge around data, it's all very well if you can find a third party provider that will give you information in a in a semi structured format. But if you have to extract information from PDF documents, from Word documents, anywhere that it's it's text based, There's a huge amount of manual work that gets done copying and pasting and retyping. And the system we've developed really uses a combination of best in class text and table extraction technologies.

27:16 And we've wrapped it into a product that allows you to upload a PDF, extract the text that you want. And then the second part of that is bringing meaning to it by allowing you to tag it. So you tag the content. The value of tagging content means you can store it in the database, and then it becomes useful. Something we find again and again with our clients is that they have a huge amount of unstructured text based data stored in many, many different formats in many different locations. And you cannot then get the value of any kind of machine learning because there's no way to to access all these disparate systems. The the the INSIG docs system allows you to structure the data by getting it out of a an unaccessible format into a an accessible database format, which you can then search across and and run machine learning models across.

28:14 And then finally, I'll I'll mention this very briefly, Exeloton. Now again, you know, it it might be surprising or it might not be surprising when I tell you this, but there are well, many of the clients that we've met run very important processes out of Excel spreadsheets. And Excel is is the best and worst tool in the world. It's phenomenal because it's very easy to use. It's phenomenal because it has ability to get things done very quickly. Where it becomes, I guess, scary is that there's a point at which it is no longer fit for purpose. And when models become that big and that sophisticated that they should be in the database, that's the point at which Excel has run out of its out of its use.

29:03 Part of what Excelerton does is evaluate the risk of a spreadsheet, how many functions and formulas it has, how many dependencies it has spread throughout that spreadsheet. But then, and I guess most importantly, it converts the Excel formulas into Python code.

29:20 Now somebody said to me, well, that doesn't sound that exciting. And I said, well, it's actually not. I said, but if we can't do it in an automated way, you have to have a human typing those out. And if you can imagine a spreadsheet of a million rows of of hundreds of tabs, that's millions of keystrokes. So again, what we're effectively selling is we're compressing a problem, which is let's get you out of this risky infrastructure into a proper database infrastructure quickly. And then the second part of that is to allow you then to carry on and get on with your business and apply machine learning and apply more interesting algorithms to your content. Yep. So I guess that's that's excellent. I think just before we we turn over to questions, it it it makes sense to to to speak just very briefly about the the the so what. So so how have how has INSIGH helped clients to optimize their performance to to increase re returns?

30:23 What what have these tools actually resulted in from a an INSIGH client perspective? And perhaps, Steve can speak to that. Yeah. So, I mean, in in each instance, you know, the this idea of modern technology or machine learning, it it can solve different problems. And it's not always about can I make more money? It's sometimes about can I can I save money or can I save time? You know, I mean, when we think of this idea of a a

30:50 an antiquated process taking a week to run from end to end, that when you lift and shift it into a scalable database infrastructure in the cloud and it runs, know, 10 times faster, and all of a sudden, something that takes five days now takes two hours, that's the kind of value that you can apply. And not only does it create headspace to think of applying your mind to do more important things,

31:15 It also allows you to get to a a wrong conclusion more quickly. And and and that's another, I guess, valuable part of this the application is that a lot of times an idea doesn't render an actual result. But But if you can get to that conclusion quickly, it makes it makes quite a big difference. You know, in terms of

31:36 some of the clients where we've looked at in terms of portfolio optimization, you know, very good strategy, you know, well placed market assumptions, well placed factor selection, how can we make them select better companies? How can we make them select better companies at the right time or get out of those companies at the right time? And, you know, we have examples where, you know, we've helped a fund outperform their benchmark by 30 percentage points. Now we can't take all the credit because they had a very good strategy, but where we can take credit is that the machine learning overlay helped them do much better. And and that's really the the the kind of space we're playing in, which is to say, you're doing something well. Let's help you make it make it better.

32:25 Just having a breath. Excellent. Why don't we, Mark, if it's all right, we're happy to turn to some questions. Absolutely. Matthew Siegman, thank you very much indeed for your presentation. Ladies and gentlemen, do please continue to submit your questions using the q and a tab situated on the right hand corner of your screen. Just simply type in your question and take, press send. But just while I let you guys take a few moments to review those questions submitted already, I'd like to remind you the recording of the presentation along with a copy of the slides and the published q and a can

32:55 be accessed via investor dashboard on the investor meet company platform. I'd also like to remind you that your feedback is important to the company, and immediately after the presentation has ended, you'll be redirected for the opportunity to provide feedback in order that the team can better understand your views and expectations. Matthew Steven, obviously, you've had some questions come through. If I could ask you just click on the q and a tab, and then read out the question where appropriate to do to do so and give you a response, that'd be fantastic. Thank you. Excellent. Well, I will, play the part of moderator and and, help Steve to answer questions where I can. But why don't I begin with the first one here? What's a typical revenue model with a client? Is it a a SaaS or a software as a service per seat model? And what's the sales process?

33:41 So yeah. So the I mean, in each of these products, and I think I'll I'll focus on portfolio in the first instance, it is a per seat model. But because it's essentially portfolio and the data product or two separate products but we sell them together, we look at a desk and we look at the usage of those two products on that desk. So we typically would like to sign up an individual user, but in most instances, it's it's five or six people on a desk. So that's really where we see us being able to do a couple of things. One, you know, fulfill the requirements of a desk by delivering the product and the data. And what we tend to see is that desks have a split of experience. So you'll have people doing the data part of it, the data collection part and then analytics.

34:34 Then we'll have the portfolio selection and then there's the kind of risk and attribution side of it. So within that, you know, we we're typically seeing that DISC is around five to six seats, which we are or selling that as a license. One of the things that I think differentiates us is that it's not us saying this is our proprietary data. We're taking the client's data into the framework and projecting it back to them through our application. So it's it's their information. It's their signals. It's their factors. It's their assumptions. And, again, the reaction that that that we observed is that they can see their portfolios come to life in our application.

35:22 And that's again, I've I've used the phrase, it's engendered a lot of trust and a lot of immediate interest because they they're drawn to it. They don't have to do any work to then make it happen. They have a system which is ready for them to use. So the goal is to to sell the the data as a service and the the portfolio product as a licensed based offering. What we see as part of our sales strategy is, and we've seen this with with existing clients, is once we've sold to an individual desk, I guess, word spreads quite quickly that there's this machine learning tool which is helping them make better decisions. The introductions to other parts of the business happens quite quickly.

36:09 And the hardest part of the sales process is the vendor onboarding, having to fill out all the security questionnaires and go through the process of getting onboarded. And the typical sales process is an initial understanding of the client's strategy, getting their data into the system, which typically takes a couple of weeks, assuming that there's no super esoteric data. It's just standard financial market data. And then they get to play with their own data in the system for a three to four week period. And then from there, it's the, you know, the contract negotiation, the license negotiation. So currently, I think we're looking at about, you know, it's it's three to four months, you know, end to end. But I guess as we as we do more of them, we will get quicker at it. And that's that's specific to the portfolio and data tools, which which require a fair amount more onboarding, whereas, the ESG tool, when that launches, will be much, it's it's much, quicker for for clients, to to be onboarded with that because it's effectively it's a web application that they'll just need a username and password for. And since it doesn't incorporate their own data, the their their access to it is is a very quick switch. Yeah. No. Yeah. That that's a really good point. And, you know, again, the the

37:40 I guess the the challenge struck opportunity in the ESG space is there's not a lot of companies that provide the central repository of the data you need to get to a decision. So in offering that, the the access is is much simpler. In the future, we will have an ability for companies to upload their own data and use our machine learning classifiers to score their own data. But in the first instance, it will simply be here is the data you need to make your decision and we'll help you get there quickly. And here's a a rather related question. What are the typical contracts in regard to length and are they per user or company wide? So we've we've we've answered the per user or or company wide, but it might be valuable to to speak about the the length and how we anticipate with the portfolio and data product in particular, the length of those contracts and how how that would work? So, I mean, it's interesting. I mean, to this stage, most of our clients are are kind of have a longer term investment strategy. So they're looking you know, they're not doing day trading. I mean, they they they have a longer term horizon. So the the contracts itself are a minimum of twelve months. And mainly because in order to see the value of a machine learning strategy when you already have a long term investment strategy, it means you have to you have to play it out.

39:01 So, you know, at at a minimum, it's twelve months. And there's I guess there's there's a couple of bits in there. You know, once the data is in the system, if you, you know, reviewing your your dataset on a quarterly basis or a monthly basis, the data framework is going to ingest that data on that periodicity and then refresh. So each time there's a refresh of data, new set of financial reports, new set of accounts, new set of numbers, you have an opportunity to run the models again, and you have an opportunity to reset and check what you're doing. The contracts we have in place at the moment have perpetual and just rolling. So we don't even have an end date at this point, mainly because part of it relates to the investment strategy, that it's going to be a three or five year investment horizon. And so we think that actually plays well into our business model because we're not on a, okay, if this doesn't perform in three months, you're out. It's actually based on the fact that, you know, the more data you have in the system, the better decisions the machine can help you make.

40:07 So, you know, I think realistically, you know, three years will be a a minimum term at which someone would may decide not to do something. But I don't know. I think that seems about right. Yeah. But I think the the experience of the INSIG PLC business thus far is that with with each of the consulting clients and and as we're onboarding product clients that that over time, you become even more and more embedded with that client strategy. And and the the the reliance that the client has on on the tools just grows. I think what's what's core as well on the portfolio and data side is that that the transformed data becomes part of the portfolio application offering or or in some instances, a a a broader off offering to the institution.

40:59 But to the extent that a client, if and when he or she wanted to end that license, then the transformed data is is no longer available to them. So so that becomes a a a a very difficult choice to walk away from from from the license. In terms of other questions, we have one. Can you explain who else competes with you and what sets you apart? So, yeah, I mean, this is a question that I actually had a lot over the last few weeks from different journalists. And, you know, it's a very interesting space because there are what I would call the mega data providers. You know, companies like Bloomberg and Capital IQ and Refinitiv, you know, they provide huge datasets, and they provide it in in in slightly different formats. Either you can get a terminal and you can pay a terminal license and you can access the information, or you can get this huge data dump or API feed.

42:03 Now they are literally at polar ends of the offering because if you use the terminal based system, you then have to do the legwork around getting that data into your format or into a spreadsheet to do to do something with it and then try and do machine learning. So there's quite a heavy lift. If you go for the option of getting all the data in a massive database, which is what you can buy, you're gonna need to hire a data engineer and data scientist, etcetera, etcetera. So in the context of the mega data providers, we fit in the middle where we're doing the transformation and the packaging and making it accessible. So I think that's that's one example of where we fit and while I think we're different to to what they're offering.

42:51 There are quite a few, I guess, smaller companies that offer specific things around signal generation, about text based classification. Do apologize there to investors. The guys have just clicked out. Give them two seconds, they'll come back in. Thank you very much. Thanks, Steven. Just follow the same process of coming through, and you'll be presented with the option just to reconnect. And there you go. Hi. Sorry about that. We had a bit problem, guys. No. You're back. You're all good. Thank you very much. Again, would you mind just what did you last hear?

43:40 I think we just we were just talking about the data providers and where you sat. I sat in the middle of the group. Okay. Perfect. Yeah. So in the context of the big data providers who who do offer services in the space, I think we offer a slightly differentiated services that we we take some of the heavy lift out of getting the data into a format that you can do something with. So that's one context. There are other providers that do much more specialists, you know, around pure portfolio optimization or pure signal generation. But, you know, we we are offering a slightly more generalized service where it's the data ingestion, the data optimization, the analytics that go with it. So, I mean, you know, we haven't come across too many people doing exactly what we're doing at this stage.

44:30 So we have a a number of other questions around the g geographic split of customers and revenue and whether or clients are mainly UK based, etcetera. So perhaps, Steve, you can take all of those. Yes. So I mean, in in you know, if if if we look at the history of what we've done from a consulting perspective, it was it was kind of a split between The US and and The UK. And the I guess as a small company at that stage, we didn't have hundreds of clients, so it really was a handful in The US and a handful in The UK. I think our emphasis now is we have obviously a strong foothold in London and we want to very aggressively grow our client base in the London asset management market.

45:20 But we also have our sales development head based in New York. And that's with purpose because we also know New York is another market that has a big asset management industry that's in a very similar position where there are a number of firms that want the acceleration of modern tech, and we think that we'll be well placed to help them. So I think, you know, in terms of what we anticipate is is a reasonably even split between, you know, The UK and and The US. But as well, I've also learned over the last few years is you also go where you get the least resistance and where the most money comes from. So we'll we'll tell you again in six months to see who's who's signed up the quickest. But I think at this stage, you know, our goal is to expand our footprint first in The UK or in London.

46:18 But again, you know, if we get a lot of interest and traction in New York, that might that might change. There's a question around how big the sales team is, and is this a key area to scale up over the over the coming months? Yeah. So it's a it's a it's a it's a good question, and it's a question that we've had a lot, which is it's it's actually a switch at the moment in our and we've been in what I would call when you start, you're in innovation and discovery mode and and hope that you find something of interest. Then you get to a point where you say, okay. We found what people want, and we've been in build mode for for some time. And now that we've been building, we're actually switching to sales mode. So we have, you know, we've hired a very senior successful sales guy recently, but the other sales supporting act is is myself, is Matt, is Yakko, you know, and as, I guess, product experts and domain experts, we're the next best salespeople.

47:19 So, yes, that's absolutely where we certainly are shifting the focus of our senior executive guys to be the the sales team. And we have a sales support function within this as well where, you know, we can obviously get in the number of doors. We have to present a compelling use case as to why we we are the right product for for a company. And then we have a a team of two at the moment, but that's we're hiring for more. That is gonna do the kind of the now you're on board. Let's get you set up. Let's get the system set up. So yeah. So it it it it is a big focus for us. And certainly over the next six months, there's gonna be a lot of activity there.

48:05 So a question here about the scale of the opportunity or or the scale of the market and whether or not this could become a billion dollar company. I did wanna just turn to slide here in our deck, which, there there there's been a fair amount of research, in recent years around, the the scale of the AI machine learning spend, particularly in the asset management sector. And I think that's that's part of what we're excited by here. Obviously, we are offering tools within this space, but tools that are in demand by asset managers. And and and and you can see that just in the the the speed of the acceleration within within the AI spend with with AI spend growing to 8,300,000,000 by the asset management industry, growing at a combined annual growth rate of of 41.

49:03 So it's it's something that there's no question that asset managers are recognizing the the importance of, and that's demonstrated by by the money that they're spending on it. And it it really comes back to my comments at at the start of this presentation around how there's asset managers and the financial services sector, generally are recognizing that this is a step change for them. This is a a point at which to survive, they need to invest in these technologies. They need to be data capable. They they they they need to have the skill sets to process large amounts of data and to apply the learnings from that data to augment their investing strategy. Yep. That's that's a good answer.

49:52 Good. Question. How have how have customers and prospects reacted to the news of now being listed on the London Stock Exchange? I see you. Yeah. No. Actually, I mean, it's it's been it's it's been quite a journey to get to the point of of of listing. And the I guess the the sheer amount of work that has to go into getting to that point is exhausting. But what I can say is the reaction has been phenomenal because, you know, the clients that we've been, you know, working with for a while are saying, wow. It's amazing to see how, you know, you guys have grown. It's amazing to see how quickly you developed in the demand. But I think more interestingly, it's it's given them a sense of comfort that we weren't gonna disappear. You know, being a startup company, there's always a risk when you're pitching for business that, you know, you're doing something for another company that's, well, what if these guys don't stick around? What if they don't loss? So there's always a bit of a concern.

50:50 And just just just the sheer number of emails and phone calls that that that we've had in the last, you know, week have been this, phew, you guys are here to stay. Oh, and by the way, we've got all these additional projects we wanna talk about. So I think the the reaction has been, you know, a lot better than I could have imagined. But, also, you know, when we published the new website and we and and we and we started to broadcast, you know, what we are we are now going to be selling, you know, specifically on the ESG side, you know, the the unsolicited phone calls we've been getting from, you know, friends in the industry we know to say, okay. Great. We're you guys are ready. Let's go. We wanna talk to you about how you can help us. So the listing has has absolutely helped us.

51:37 And it's I mean, obviously, it would have raised the profile and, you know, there's more press and coverage about it. But I think more importantly is that we're no longer a small business that given unfortunate circumstances can disappear. We're actually a business that is well capitalized. It's got a lot of support, and we've also massaged some incredible people. So I think it gives other companies more comfort that we're here for the long term. So question. Are you seeing clients ask for other data from other areas other than ESG such as crypto? So it's it's interesting. Like, machine learning AI and crypto all inter intertwined.

52:24 We have. And it's you know, we we've actually spoken to a few crypto companies that that share common platforms. And, you know, the the I made a comment earlier around, you know, there's a common theme across all of this stuff is that with all of the interesting process and analytics, you have to do the boring stuff well. So the interest around crypto is is is also around, well, how do we make sure that we can process and analyze information effectively and on a repeated basis? So we have actually had a few inquiries around companies where they're asking us to us to help them from a infrastructure perspective. I think we're gonna get more.

53:10 But yeah. I mean, I think that's next next week's Challenge. Challenge. Yeah. Yeah. Another one. Who are the major shareholders of the company, and how are management incentivized? I'll I'll begin with an answer on that. Yep. So the there are a number of of historic institutional and and and other shareholders of Catena Group. But at the point that this transaction completes, pending shareholder approval next week, the primary shareholders of the company will effectively be the cofounders of of INSIG Capital and some of INSIG's shareholders. So management is very well incentivized to meet targets and to grow the business rapidly.

54:01 Yeah. I mean, guess just another point on that. I mean, one of the fundamental beliefs we have is that the people that work in the company need to feel that they have ownership in the company. And all of our full time team have shares in the business. And being part of a listed vehicle means that those shares actually mean something. And and I and I say this with a with a smile, you know, having worked in in Silicon Valley, you know, part of the negotiation is we'll give you very little salary, but we give a lot of share options because we're gonna become a unicorn. And, you know, as as my my start ups don't become anything. So the fact that we can actually give people a shareholding in this business is very important because the better we do and the better they do in their job, you know, it's going to have a direct impact on their own lives.

54:57 Another question. When will your latest figures be announced? So the latest Catena figures will be announced by September and those will be as we change our accounting reference date, those will be fifteen month figures to a March 31 accounting period. But as the the INSIG Capital acquisition was not done by that period, the INSIG Capital figures wouldn't be consolidated in that. However, I'd I'd I can't commit that with with the appropriate support of of our of our broker, Nomad, we will be providing as regular trading updates as as we can over the course of of the coming months.

55:46 That's fantastic. Matthew, Steven, thank you so much. You've answered every single question. So that's wonderful. Thank you. And if there are any further questions that do come through, obviously, Matthew and Steven will be able to review those post event. Perhaps I can just ask you both just for a a final few closing words just before we redirect, the attendees to give you guys some feedback. I'll I'll just say very quickly, thank you very much for taking the time to to listen to us today and to to ask some some questions. For for the new investors who've joined the call, very much excited to have you on board and excited to continue the dialogue with you during the course of of the coming months. I I know I I speak for Steven saying that that we are, as as the the the new insight INSIG AI team, very excited about the coming months. I know that the the broader INSIG AI team is is excited about the future for us. So thank you for taking the time today, we we look forward to having further conversations with you.

56:44 Yeah. So I I mean, I echo that. You know, I mean, I guess, yeah, our our belief is if we continue to focus on building, you know, solid software with exciting technologies, you know, the the share price should follow. You know, we've we've we think we've proved it with our three year history with our current client base. And our goal is to continue doing that, but we're just going to do more of it and and quicker. So, yeah, so let's watch the space. Fantastic. Matthew, Steven, thank you very much indeed for updating investors today. Could I please ask investors not to close this session as you will be automatically redirected for the opportunity to provide your feedback? If you've accessed the meeting from the website, your feedback will appear in front of you. If you've accessed the meeting via the link sent to you in the email, you'll simply be asked just to log back in. It'd take just a couple of moments to do so, and it'd be greatly appreciated by the company.

57:36 On behalf of Catena Group PLC, we'd like to thank you very much for attending today's presentation. That concludes today's session. Thank you, and good afternoon.