Good afternoon, everyone, and welcome to the Cerillion interim results webinar. Presenting today we have Louis Hall, CEO, and Greg Price, CFO. All other participants will be in listen-only mode. You are welcome to submit written questions either at the end or during the presentation using the Q&A tab at the bottom of your screen. Management will attempt to answer all of your questions at the end of today's presentation. Thank you for joining today's call. I'll now pass you over to Louis Hall, CEO. Louis, please go ahead. Thank you, Gareth, and good afternoon, everybody. Welcome to the webinar. Just introducing ourselves, I'm Louis Hall, CEO of Cerillion, founded the company back in 1999, a very long time ago, as an MBO from what was then a large U.K. software house called Logica. We then IPO-ed the business finally in 2016 and have seen a pretty successful run on AIM since then. Greg has been with us four weeks, so bear with us because he's still getting up to speed, but done a terrific job this week in what's been quite an exhausting roadshow, and he's already an old hand. Greg was previously at essensys, another AIM-listed business, and also at Checkit, another AIM-listed business where he was also CFO. We'll move into the meat of it. Highlights for the half. One of the key points we're making is that we're very pleased with the doubling of new orders from around GBP 40 million for the same period last year from GBP 20 million, the previous first half, to around GBP 40 million for this half year. That's a significant jump. It's a new record. That was largely facilitated by closing the large deal with Omantel that we announced in January. That's a significant deal for us, not just in terms of setting a new bar in terms of scale. It's almost double our last largest ever contract win. It's also a beacon for other telcos in the Middle East who traditionally buy from very large brands. For a large Middle East telco to buy from Cerillion is really helpful, and that should give us more traction in that region. Got to bear in mind that Omantel are the bigger shareholder in Zain Group, which is another regional telco which has, I think, eight properties across the region, eight other telcos across the region serving 50 million customers, so that's a lot of potential for us to go after as well. A very important win, not just in terms of global scale and credibility we have from a much larger telco, but also regional resonance. All of that's pushed the back order up 64% to a record GBP 82 million. Obviously, that gives a lot of confidence looking into H2. I think the other thing that we're quite pleased about is that the new customer pipeline, even after taking out the significant value of Omantel, actually increased a bit from GBP 261 million to GBP 271 million. That gives us confidence that we can replace the pipeline quickly because there is strong demand in the market. There is a disparity between H1 and H2, which we flagged at the trading update. Revenue is actually down compared to the prior period, and that plays through into margins, obviously. It has a sort of disproportionate effect on margins because the shortfall in revenue in the first half, the largest part of that is license revenue, where a bit like last year, we expect most of this year's license revenue to be recognized in the second half rather than the first half. License revenue is much more profitable than services or other revenue. We feel confident that we have the backlog to deliver the H2 numbers and if we thought we weren't going to achieve that, we would have said something. I think you can assume that we fully expect to achieve those numbers. Consequently, we've pushed the dividend up by 15%, largely on the back of that confidence. In terms of operations, we've been working very hard on our new implementation at Ucom in Armenia, a deal we closed last January 25. It took a bit longer to get started than we'd hoped, but it is now through most of the initial delivery phases, and we're moving into sort of final testing and network integration on them. Cutover data migration, we expect to happen in the autumn. It's important to get that customer live there because there are other opportunities in the region that we'll be looking at this one and having that successfully launched will resonate. There's been some concern, I guess, notably people are concerned about the Middle East crisis and war with Iran and what the impact that is, what impact that's having. What we're seeing is not really much impact. We've had people on the ground there since we started this work in February. There are people there now. The planes are going backwards and forwards every day to London still. However, if we couldn't keep people on the ground and we did have travel restrictions, we could do all this remotely, just as we did during the pandemic. We grew faster than ever during the pandemic, didn't have anybody on the ground anywhere. We don't really see that as a problem. That project's progressing well. We've pretty much finished the requirements phase, the hard part where we define exactly what we're going to do. We're about to move into the heavy lift of configuration and so on. That's going well. On the R&D front, we released our, in the first half, 26.1, our latest product release. The focus there really was around agentic AI and more functionality across that piece. This is things, say examples like supporting customers to choose products, supporting customers to understand their bills, which sounds, I guess a little bit, is that really interesting? I think the important thing is that for telcos, it enables them to offer much better customer service to their customers. We're not just talking about chatbots here, we're talking about dialogue with these intelligent agents. Being able to say, for example, "I'm moving to Oman. I'm going to be in Oman for nine months of the year. I'm going to be back in London for two months, I'm going to be Australia for Christmas. By the way, I like sports. I want to watch some sports channels. Oh, hang on, family want mobiles. Few of those, please. You're then in a dialogue with the agent. What kind of handset would you like? Is it Apple? Is it Samsung? What kind of sport do you like? Baseball, football? This is really quite incredible stuff. Of course that's generally better service than you'll get from anyone but the most advanced, experienced customer service agent as a person, and it's not a person. You save on the cost of the person, you also make it easier for the customer to onboard onto your platform. For telcos, this is really important stuff. That was the main focus of R&D in that release. In terms of general position and sort of overview, nothing fundamental's changed in that we still have our main bases in the U.K., India, and more recently, Sofia in Bulgaria. In terms of the revenue mix, this in the first half software was a bit lighter than it would normally be because of the relatively low license revenue recognition. We'd expect that to adjust more to a higher percent of software and services in the second half and in the full-year outcome. As ever, a lot of our businesses in Europe, as you can see from this chart down here. I think we'll see an increase in Middle East and Africa, obviously with the Omantel project getting going in the second half. As ever, we have a lot of customers been with us for a long time, and the majority of our revenue in any given year comes from customers who are already customers at the start of the year and have been customers for at least 12 months before the start of the year. That is an important factor in terms of visibility and the fact that we haven't got to keep winning new logos all the time to achieve our numbers in any given year. Just a little bit for those of you who are new to the story, I appreciate. Apologies those of you who know the company already. For those of you who are new, what do we do? An important question. We provide enterprise software that acts as the glue between telecoms businesses, network infrastructure, and their customers. It's the software that enables telcos to define the products that they're going to sell. Those are often quite complex bundles of TV, broadband, mobile, fixed wire, whatever. You have all kinds of cross-subsidy discounts here, there, and everywhere. That's a complex thing to define, and most telcos offer hundreds if not thousands of these different combinations. All of that we have to model and manage in our product catalog. Once we've defined those products, then we provide the software that enables telcos to onboard those customers. We spoke about agentic AI, but we also support traditional CRM, traditional self-service mobile apps and so on. Those are all still very much the bread and butter of telco business. Once customers are onboarded, we have the software that manages connection of those services on the network, which we call provisioning, and that's a complex set of workflows that have to happen in certain sequences. Once customers are enabled, they're live, then we have the real-time software that manages their usage. Manages their balances. Can this customer connect this internet session? Have they got enough data balance? Have they got enough call time balance, et cetera? That side of it, at the end of a billing period, we're billing customers, we're collecting payments, we're handling credit control and so on. A vast array of different functionality that really is the core of what telcos do. We sell this through a suite of modules that address different areas of that piece. I won't go into them all in detail. We haven't got the time to get into all that level of detail. What is worth saying is that most customers generally buy most of the modules, because one of our key differentiators is that these modules are built to work together from the start. If you buy modules from different vendors, you've got the challenge of having to integrate them and make them work together. Ours already work together, generally, most customers will buy most modules. Occasionally customers will buy a single module. For example, they might just buy the product catalog, or they might just buy the Converged Charging Platform. Of course, we've got the opportunity to upsell other modules later on. In terms of differentiation, our key differentiators are that we are a product solution. This industry has been dominated by very large vendors who typically deliver heavily bespoke versions of their systems. That is a much more services-heavy, higher risk, longer duration, higher cost exercise. We're starting on day one with a product solution that works, so we can demonstrate real software straight away. We haven't got to bespoke it or tailor it or build it. It's here. Customers can start literally in meetings where we're defining requirements, looking at real software, and that's quite powerful. Of course, it means that our total cost of ownership for a customer is significantly lower. Our time to market is much faster. Typically, we're delivering these systems and getting them running in 12 - 18 months. That might sound a long time, but given the complexities of configuration and data migration and so on, complexity of platforms, that's actually pretty quick. Whereas the bigger bespoke systems, bigger vendors with their bespoke solutions, will take three to five years generally to do the same thing. Of course, the solution you have at the end of all this is much more flexible, both in terms of ease of use and the fact that it's designed to be used by our customers, who don't require experienced software engineers to amend products or introduce new workflows. The customers can do all that themselves through configuration, also in terms of upgrade. It's important these days to stay up to date with the latest security patches and so on. If you have a bespoke solution, all that has to be done specifically for that customer every time, which is very expensive. Whereas with our product kernel approach, we're just rolling standard releases out to customers as we go along. We provide the solution, typically these days in the cloud, mostly private cloud, because most telcos won't put their data in the public cloud. Once the system's implemented, we support that system through standard support and maintenance. We're typically hosting the solution for the customer in the private cloud, and we're operating it for them as well. We're running managed services. Essentially, the customer is consuming the solution on a SaaS basis, and that shapes the way we sell it. We sell five-year subscription agreements. A customer signs up to a subscription fee based on the number of end customers they have. If a telco has 1 million mobile customers, they pay 1 million times X a year. If they have 2 million, it's 2 million times X a year, and so on and so on. In addition to that, they will buy a fixed price implementation project to put that software into use. That's the sort of 12-18-month project that I'm talking about. A very broad range of customers across most geographies. We have some logos you'll have heard of, like Virgin Media, Orange, Airtel, Liberty Global, et cetera. We also have some smaller telcos. Obviously, we've evolved over quite a long period, starting off with some very small telcos to now work with some of the biggest names in the industry. I think one of the key evolutions we've seen since IPO in 2016 is the gradual increase in scale of customer and scale of engagement. If you look at when we IPO'd, a typical good deal for us would have been about GBP 3 million worth of total value, up to the point today we're at GBP 42 million. That's very important because the larger customers that we're more deeply engaged with drive a lot more upsell, a lot more account business. Our model does rely on customers carrying on spending, whether that's upgrades or license expansions because they've grown or acquired other businesses or buy new modules or general kind of consultancy, training, whatever. Those customers do spend heavily, and the larger customers spend very heavily. The more of these larger customers with deeper engagements that we can win, the easier that it becomes to grow and to keep that growth moving. In terms of market size, these are figures from last year from a telco consultancy called IDC. Their forecast is that annual BSS/OSS market revenue will grow to around GBP 60 billion by 2029. I guess they're in the sort of mid-GBP 50s billion for 2026, 2027. The important thing is that it's a very large market. We don't address all that market. For example, we don't sell to China for obvious reasons. Even if you take out those sorts of things and said we can only address half that market, which is a bit of an extreme statement, but if we are only addressing half of that market, it's still a huge market, and we still have a very tiny share of it today. There's a huge space for us to grow into, which gives us a lot of comfort. There are high barriers to entry. Despite the talk about AI and displacement and so on, something we try to address in the RNS, but essentially it would be very hard for a new entrant to build all of this from scratch with AI. With AI, you still need to tell it what to do. The problem is that with this kind of software, it has to be 100% accurate all the time. If you don't send the right commands and the right order to the network, you won't turn the services on. You can't send a customer a bill that's 95% right. That's not the way it works. To get that level of precision, the amount of detail you'd have to go into in AI prompting is vast. In the end, all you're really doing is instead of writing Java code, for example, you're writing prompt code because it's that level of detail you've got to go into. Once you've actually done that, in many ways, it's not much different to building a new solution from scratch in the traditional way. You get some efficiencies in testing and so on, yes, but you've still got a massive process to go through. Once you've done that, you've then got to get a customer to adopt it. Telcos incredibly conservative and very reluctant to buy a solution that's not already proven in the market. If you could get one of those customers on board, which would probably take you a year at least, then that's got to be implemented, which in most cases is going to take another couple of years. You've got to see some proof. We're talking four or five years or more from a standing start, even with AI, even assuming you can build this with AI successfully, to get something that becomes a competitor for us. We're really not seeing any new entrants trying to do this. If you're an existing vendor, you've got enormous advantage of being able to use AI tools to improve what you already have. The values and the knowledge that's in the software, the knowledge that's in the business logic in the software, if you're already a vendor that has all of that, then you can improve a solution that already has the embedded knowledge from that position. Things like architecture changes, technology upgrades, they become vastly simpler with the AI tools. Of course, we're using those AI tools to do those kinds of things. I appreciate we're a bit tight on time, I'll keep moving. Talking too much as usual. Just very quickly on the competitive landscape, because we have seen quite a lot of change here recently. These are just some examples of transactions that have gone on in that landscape. CSG is probably our number four competitor, acquired by Netcracker, which is part of NEC out of Japan, which is probably our number 2 competitor. That's reduced the shortlist on RFPs by one. I guess the top four is now a top three, in terms of the people we compete with. Optiva, which was historically quite a strong competitor, finally got acquired by Qvantel, which is a relatively small Finnish player. I think Optiva, sorry, not obvious to everybody, but Optiva had hit the buffers and been in quite a bit of trouble for a long time. Qvantel is a relatively small company, probably similar sort of size to us, so whether they've got the resources to turn that around. There has been no investment there for a long time. It's hard to see whether that's going to happen, but maybe it will. MATRIXX was the last surviving standalone charging platform. That's now been acquired by Amdocs. That again reduced the number of options that telcos have to go to with their RFPs. Finally, Comarch a little while ago was taken private by CVC Private Equity. And they were quite a strong competitor prior to that. We're seeing less from them since that's happened, and I think there's a lot of shuffling going on there to decide what bits of that business to keep. It's part of a larger conglomerate that CVC have bought. There's some disruption there, I think. Again, I'm sort of conscious of time, so I will keep going. Greg, would you like to say a few words about the KPIs? Yeah. Thanks, Louis. I will. Hi, everyone, and good to speak with you. I'm very pleased to report my first set of results as Cerillion CFO. While there is a significant weighting to this year's results, as Louis has said, we believe that we're well-placed to deliver on market expectations for the full year. In terms of our key KPIs, revenue was down 14% to GBP 18 million. This reflected the phasing of new orders with minimal software license revenue recognized in H1. The impact of this also is evident in adjusted PBT, which was down 41% at GBP 5.5 million, and on our EBITDA margins as well at 34.5%. However, the underlying health of the business remains robust. If we look at the bottom line of this chart, the recurring and annualized term license revenue, which is adjusted to show term licenses on an annualized basis rather than upfront when customers have the right to use the licenses in line with IFRS 15, that shows growth of 5% to GBP 19.1 million. That's probably a good metric to consider. In terms of cash, our balance sheet is still really strong. We have net cash growing year on year by 4%, which is then driving the dividend increase that we have. A quick run through there of the KPIs. In the interest of time, there are more slides on the financials that I can go into to cover the P&L, the balance sheet, and cash in more detail. For now, I'll hand back to Louis, and we can cover any questions that you have later. Yes. Okay. Just dash through these in case it triggers any thoughts. Okay. Finally, a look at pipeline orders backlog. As I said at the beginning, a significant jump in new orders, which is really important in terms of visibility for H2, but also looking into 2027. Despite that, as I said earlier on the pipeline, the prospective customer pipeline, this doesn't include existing customer business. We have a separate pipeline which we don't disclose in this presentation for existing customers. The total prospects value is higher than this. The weighted value, that's after we adjust for percentage probability across all the prospects has remained fairly robust. Of course, that's pushed our back order up significantly to this new record of GBP 82 million, which you can see is quite a big step change. All of that gives us confidence. Just a quick summary. Omantel is a really significant change proof point, not just with other larger telcos, but across the region. We've spoken about the H1 waiting, H1, H2 waiting, and why we think that isn't a problem, although it is a little unusual, we appreciate. We spoke about new customer pipeline just on the previous slide, of course, the balance sheet remains very strong and cash has grown again by about 4% as Greg alluded to, I think. That obviously gives a lot of comfort. I think, generally speaking, we're well positioned to achieve expectations and I look forward to doing a lot more, not just for the rest of this year, but to the next few years ahead based around our overall ambition to grow, to double the size of the business again in the next 3-5 years. That's the end of our presentation. Back to you, Gareth. That's great. Thank you very much to Louis and to Greg. Overall, very good. I will now open up to questions from the audience. If you would like to ask a question as a reminder, please click on the Q&A button at the base of your screen and type in a question for me to ask on your behalf. We have already had quite a few questions submitted, so I'll start running through those. First one, Louis, is can you please discuss the risk of customer concentration or, I guess, loss of customers based on the concentration? We have very little churn. Once the customers are on board, it's a big process to move. Typically a telco will spend 18 months, 12 months, 18 months, maybe even a couple of years choosing a new solution. Once you've done that, then as I was saying with us, it's 12 to 18 months. With most of the competitors, 3 to 5 years. It's a long, long process. If you're not looking to move almost as soon as you've started or as soon as you've gone live with us, then you are going to achieve that in time to hit the end of the initial five-year term. Obviously, it's important that we have to do a good job. We have to maintain strong customer relationships. It doesn't fall off a log, and that is a big part of what we do on an ongoing basis. Customer concentration, when a customer is a new logo, a new customer, and we're doing all that services revenue, typically in the first year or year and a half that will have a significant concentration impact on that financial year. Of course, once that project's gone live, that services part drops out, and that customer becomes a lot less significant in terms of revenue concentration. The same thing happens with a license revenue recognition. Obviously, that has a big impact on one year's revenue, but it doesn't then recur. Customers can be quite large, heavily weighted in our mix in the first year, but then not in subsequent years, but they become kind of more normal. Does that make sense? Okay. That's great. Thank you. We've got a few questions actually on the subject of Oman. I might go through those one by one, but I'll sort of cluster them together a wee bit. The first one is what are the major challenges in terms of payment conditions or cash collection and implementations in a Middle East customer like Omantel? Cash, I think there's a lot of bureaucracy, but fundamentally, this is a company that is very well-funded and can pay. Of course, ultimately, if they don't pay, there's always the ultimate sanction that we can turn it off, which means that they're out of business. Not that we would ever use that, but that's always there in the background and something we could use. We've been paid in some very obscure places in the past, and we always get paid. I'm not really too concerned about that. There is bureaucracy, and it is quite complicated. I think that also goes for implementing. Implementing in Oman, it's a professional organization. I don't think there's any particular reason why it'll be any more difficult than anywhere else. Okay. The next one on that subject is, what was it in particular about your offering that won you the Oman deal? I think the customer liked the product-centric model. They liked the fact that they can connect to our APIs to do other things around the edges, for the edge cases. I think we were significantly cheaper than the other options. I think the fact we could, for example, demonstrate AI features was important, whereas the others weren't able to do that. It doesn't mean they won't be able to do that in the future. Of course, they're all saying they can do that, but if you're in a race, the fact you've got it there today I think is helpful. We had a good sales team that knew how to get around the organization. There's a number of factors, I think, but I think fundamentally, they chose the product-centric approach with all those advantages, the lower TCO, faster time to market, more flexible solution. Designed not to need an army of people from the vendor on-site all the time, which is what a lot of the competitors would require. Our approach is we implement and we're gone, and then we operate the solution for you, but you don't need an army of people on-site. I think all those sorts of things made a big difference. Okay. Thank you. A sort of related question on geography. Question is, looking at the MEA region performance, obviously you've won Omantel. You've described their holding of the Zain Group. Do you have other significant pipeline opportunities in the Middle East and Africa? We have some. I think we'll have a lot more relatively soon as Oman progresses, and we get the chance to build that out story. That's not a particularly strong weight in our pipeline right now. Okay. Thank you. Next question is, have you noticed any real efficiency from AI when making new implementations? Yeah. The thing about implementation is a lot of it is communicating. A lot of the information we need is not necessarily written down anywhere or in any data system. A lot of it is talking to people. How do you interact with that department so we can build that workflow? How do you provision that particular service on the network, and what sequence do you do those steps in? Some of that is in systems, in data, some of it's written down, but a lot of it isn't. A lot of the work is that interviewing and collating information. When it comes to the kind of spreadsheet stuff, that obviously we can automate. A lot of data migration tasks, which are quite labor-intensive, in the past, we can use AI to expedite. It's making a difference. It will remain a significant task. We also had a related question saying, have you noticed any tangible R&D savings related to the use of AI in product development? Absolutely. Yeah, we're kind of fully engaging with these tools, and of course, they're evolving so fast, but at the moment, we're making quite a lot of use of Claude, for example. That may well change when it's usurped by something else in another six months or whatever. We are getting savings in that nobody in Cerillion is writing a new line of code themselves. The tools are used to generate code, but they're generating it at a function level. Take this parameter, do this, and output these parameters. We're not saying build a new module. The level of detail required is too great for that. Where it's really helping is in the testing process. What we can use AI for is to look at a function and build a test suite and then run those tests. When the tests have been run, tell us what the errors are, what's caused the failure and why has that test failed, and it will then actually go in there and change the code. That is a big saving because about half, or at least half of the effort on software development is generally testing and fixing. Only about half the effort of software development is actually writing stuff. If you can save a big chunk of that half, then that's a massive productivity saving. Okay, thank you. We've got a question about people, actually. How do you incentivize local employees in India and make them feel part of the Cerillion culture? Would it be fair to assume that the employee turnover rate is higher in India versus other places like Bulgaria or the U.K.? That's a really good question. Actually, we've got really low churn in India. I think it's probably only about 5% or 6%, maybe a little bit more, 7%, but it varies a bit from year to year. Amongst the senior people who've been with us for a long time, it's very small. It's about 3%, I think. We don't have big churn. I think what we offer in India is the ability for people to make a difference, to be engaged in real cutting-edge R&D, whether it's agentic AI or whatever else it is, real-time charging platforms, and to own things. That's quite different to most Indian software services businesses. There isn't much product business in India. That's also quite unique. Most of the software work in India is services, and if you're a member of the five teams that are working on one part of a function for Citibank or something, you've got very little engagement with the front, the users, the customer, and you know what the whole purpose is. Whereas if you're working with us, you're very much engaged with customers, with the product salespeople in Cerillion, et cetera. You're working on projects with customers where you're seeing real difference being made, and you're delivering systems that are changing their business. I think we have the advantage of being a smaller firm. It's counterintuitive because, of course, the conventional wisdom is that people in India generally are drawn to the big brands and the big firms. I think increasingly people are finding the big firms are just a bit soulless and they are commodities. Okay, thank you. Next question relates to the order book and performance. Saying, while the doubled order book is good, profit before tax is down 41%. What can you say to reassure investors both for H2 and particularly beyond that? Also, what is the FY 2026 expectation for profit before tax? Yeah, sure. Shall I answer that, Louis? Please. Yeah. Yeah. In terms of the FY 2026 profit before tax, consensus from the analysts is for GBP 22.2 million. I'd point to that. In terms of how we look at the second half, I guess the best thing I can say is that if you think of our first half as a foundation, so we had revenue of GBP 18 million in the first half. On top of that, we can expect to recognize license revenue on the Omantel deal in the second half. As I said earlier, we recognize very little software revenue relating to licenses in the first half, so that will be quite a shift. Also, the profile of our services, of our projects means that we expect to recognize more service revenue in the second half. Finally, there are also existing customers where we are talking about renewals at the moment. We expect more license revenue relating to renewals in the second half. When you add all of those things together, that's where we get our confidence relating to the second half. Okay. That's great, thank you. Next question is, what's stopping you currently from considering acquisitions as a way to accelerate geographical expansion and expand into new customers? Yeah. We do look at acquisitions, but we look at bolt-on opportunities where we can buy product that is adjacent. Businesses that have product which is adjacent to the modules that we offer so that we can upsell that new product to our existing base, but also with that acquisition, bring in other telco customers. Typically, niche players will have more mid-size or tier 1 logos than across the board provider like ourselves. Then have the opportunity to cross sell, upsell into those new customers we're bringing on board. I think just acquiring the geographic expansion doesn't really help much. The telecoms market is completely globalized. The same solutions work in any telco in any country, because it's a completely standards-based business. For us to buy another billing vendor, say in the U.S., it would give us some U.S. customers immediately, but it would give us the problem of two overlapping product lines, and how we ever merge those, and so on. I think that's just sort of not really practical in terms of trying to keep growing at this rate. Probably slow us down rather than speed us up. Okay. Thank you. We've got a couple of questions around pricing. First bit is, do you follow a similar pricing strategy approach for all customers, or do you segment them by budget and value add? Sort of aligned to that, are business development managers or salespeople restricted from offering discounts even in tough negotiations? Which I guess sort of goes together. Yeah. Sorry, I'm trying to think. I've lost the track on the first one. What was the first one again? Sorry, the first one was, do you follow a similar pricing strategy approach for all customers or segment them by budget and value add? Yeah. We do have geographical regional pricing on the subscription fees. In terms of, we don't generally tailor pricing to budget. If a customer doesn't have the budget for the particular thing that they're tendering for, we can often tweak the scope so we can look at reducing the scope of what we're providing to hit the budget. We can work share where the customer takes on a chunk of the implementation work, for example, so that reduces the cost. In terms of discounting, so it's not like we're selling vacuum cleaners and the sales process is long and it's a team effort. Lots of people are involved and there are lots of review steps. It's not the case that there's one sales guy going to the doorstep and he's got to close a deal that afternoon. There are lots of rounds of this. We'll typically respond to an RFI with some headline numbers, and then there'll be more specific criteria in an RFP response. All the time we're getting guidance, through our channels as to where the budget is and what the pricing needs to be. The customer might then come back and say, "Yes, fantastic, we need 10% off." There's a discussion. It's not the sales guys don't need to be empowered to be able to give anybody a discount. The deals are such a size that there's a full engagement from management all the way through. Okay. Thank you. Next question actually is going back to AI. It's a different tack, though. How do you factor AI related product features into your pricing? That's a good question. The way we look at this is that we are charging for consumption of APIs. If you think about AI agents, our policy is not to lock out other agents and other vendors, but to have a sort of open wall policy where our customers can use our agents, in which case they're paying for essentially each time they activate that agent, well, each time that agent communicates to our APIs, through our MCP Server architecture, which won't go to all the detail on, but it's essentially a gateway that enables our APIs or external APIs to connect into our environment, access our API, their business logic and data and so on. That's done on a consumption basis. If a customer's own AI agent or third-party AI agent wants to come across and be plugged in, we charge a fee for that access as well. Whether you're using our agents or anybody else's agents, you're still going to pay a consumption fee. That's quite different to the main platform model, which is based around number of subscribers. The point here is, of course, that if you have 1 million subscribers as a telco, maybe only 100,000 of those are going to use the AI agents to start with, in which case, paying a percentage of the overall platform fees it's hard to manage. Also, if this takes off as we expect it to in a consumption-based charging model, should be very lucrative. We're kind of then aligning with what the AI providers, the AI LLM providers are doing. They're operating a very similar model. Great. Thank you. Next question is: how do you differentiate yourselves from local competitors to attract and retain the best talent? Again, it's going back to what I was saying about India. I think if you're a product business offering opportunity to build new technology and be engaged in some of the leading-edge technologies of the day, then it's an attractive place to be. We don't pay crazy salaries, but we pay at least the top end of the market salaries. I think it's mainly around the work we have to give people and the opportunities they have to not just do interesting work, but also to move on quickly. If you're ambitious and want to progress your career, we have a pretty flat structure, and we encourage people to push themselves, and I think that's also important, especially with younger people. Okay. Thank you. Next question is: Are you actively considering a, it says FTSE, I guess it means LSE main market listing? What would be the advantage of listing there relative to AIM? Well, I think we do look at this from time to time. AIM has been very good to us, really. I think we're not a company that's been marooned on AIM. If you look at where we started out at GBP 0.76 a share in 2016, we can't really complain too much, even though the last couple of days have been a bit of a bruising. I think there's a little bit of a concern that we don't want to be the last tech stock on AIM, and it looked at one time that was maybe where it was going. I think the advantage you get with moving to AIM are really around the top of my head, I can't remember the proper term for it, but automated trading, there will be more liquidity potentially, there will be more trading, and that will benefit everyone, in theory. The ETFs I am talking about, those are obviously more focused on main market, particularly FTSE 250. I think, unless we are confident we would be in the FTSE 250 and stay in it, we wouldn't make that move. We are still a little bit away from that. Okay. Thank you. Next question is, can you speak about potential new logo wins and the size of those potential wins versus Omantel, and whether or not any one or more of those are needed to meet consensus, or could that lead to upgrades to numbers? Yeah. We're not relying on any new logo wins really to do the 2026 numbers because we're quite late in the year now, and there's only a limited impact a new logo win would have at this stage. For 2027, it's important. There are some opportunities in the pipeline that are same scale or larger than Omantel. I think if we closed one of those, then we probably would be looking at upgrades for 2027. Yeah. Okay. Thank you. Next question: Is services as a percentage of revenue expected to decline over the next three years? If so, does this mean a rise in margins given the higher software mix? The answer to that is yes and yes, in simple terms. As the larger customers with bigger licenses grow and the license book expands, the proportion of services, a lot of which, I mean, a lot of it is derived existing customers, but a lot of services comes from new customer implementations. That proportionally will decline, hence the mix will shift more towards software, and that will inevitably push margins up over the medium to long term. Okay. Thank you. Next question: have you noticed any risk of potentially customers starting to insource their software development? Yeah. That's what we touched on earlier on when we spoke about AI development. No is a simple answer. The thing is that there's a lot of talk about this insourcing and so on, but you have to look at the context where a lot of businesses, particularly telcos, for the last couple of decades have been trying to outsource as much as possible. They just don't have the people to do this stuff. AI is not magic. It requires people to direct it very specifically to do what you want it to do. That requires technical people, whether they can program in Java or whatever, it still needs technical input. People who understand the real detail of the complexity of networks and all of that. If telcos don't have those people to the same extent, then insourcing means not just doing it yourself, it means finding people who can do it, because most telcos just don't have these people. Okay. Understood. Thank you. Another question, actually, in terms of the listings. We're looking backwards. The question is, looking backwards, is there anything that you would've done differently nowadays compared to when you listed the company? That's a good question. What we've done differently. I think if you're not so much when we listed the company, because we've been going for 15 years by then. If you go back to 1999 and 2000, the heady days of dot-com boom, I think there are definitely some things we would've done differently to what we knew then. The concept of raising private equity funding and there was a sort of real focus on just spending all the money as fast as you can, hiring a big team, the devil take the high road, and if you create the team and burn the investment, then the sales will come. Of course, that's not what really happens. I think we would've taken a more cautious approach. For example, back in 2000, 2001, if we were our older selves. We traded through the dot-com collapse, and we didn't raise any new money again. In the IPO, the money we raised paid off the PE funds, and it wasn't to bolster the business. I think we got through it, but we made it harder than it needed to have been. Okay. Thank you. A question about some of the competitive landscape. You mentioned numbers four and number two players in the sector. Which number is Cerillion? Well, number one, of course. Well, what I said was not so much the number of position in the market, more people we compete with the most. My ranking is not where they sit in the market necessarily, but where they are relative to competition to us. Amdocs is the biggest competitor we have in terms of number of times we come across them and number of times we beat them. Netcracker's probably second after that. Another question about a competitor, actually. The question is, why do you no longer mention Hansen Technologies as a competitor? What's your view on their telecom-related products? We don't really see them anymore, is the simple answer. I think the Hansen model's quite different. They're an acquisition roll-up vehicle, and they're constantly buying new businesses in this space somewhere, but tend to then focus on cutting out costs, typically sales cost, not really too focused on winning new logos. I'm trying to be polite here. Making the most of existing customer base, shall we say. It's a very different model. It's been very successful for them, but it's a very different model. Hence, they're not as active in terms of new customer acquisition. I think that's probably why we don't see them. That's my take on it, at least. Okay. We've just got a last couple of questions given the time. One question, what is your current position in North America, and is this an attractive market for Cerillion? It's a market that we haven't exploited enough, and I think there's a lot of opportunity there for us to exploit it more. We've recently hired another key lead resource in North America to head up that and build more prospects there. We're going to be recruiting more pre-sales people there. It is an opportunity for us that is yet to be fully exploited. Okay. I think we'll make this the last question, given the time. Just looking at five-year license renewals, how does the renewal license for those customers compare with the original license cost? Essentially the way it works is that we agree a subscription fee and at the start of essentially when we sign the contract, and that fee then escalates with an inflation index every year through the term. When we get to renewing the term, we generally just roll another year of inflation. We're not trying to be smart and doubling the price or whatever because these are long-term relationships, and if you do that, it resonates around the market and the next customer isn't going to select you. You've got to be quite careful about how you manage that. We do it in a kind of transparent way. It's just kind of a given that costs go up, therefore, we're going to increase these fees with inflation, but we're not going to gouge you. The other thing to say, of course, is that the fees go up as customers grow because as they increase their subscriber numbers, we increase the subscription fees, and that's all programmed into the contract as well. It hasn't got to be negotiated. The fact that you haven't got to negotiate really on the renewal, it's actually helpful as it means it's a more seamless process. Whereas if it had to be negotiated every time, it would just take longer and be more time-consuming, harder to get done. Okay. Well, that's great. That's a nice note on which to end. We have now unfortunately run out of time, so there's no time for further questions. I'll just hand back to you, Louis, for any final closing remarks. Well, I just want to say thank you all for listening. Appreciate your time. We've come a long way. I think we've got a long way to go, and we're very confident about the future and seeing our product-centric model continue to get adopted by larger and larger telcos in the market. That's great. Thank you very much, both. Thank you all for attending. This is the end of the webinar.