All right, next up, we have Endava, a digital engineering provider with over 11,000 global professionals. Its heritage base of operations remains in Central Europe, but the company's increasingly going global, including more nascent, growing base in APAC through inorganic and organic means. With us from Endava is Mark Thurston, Endava's CFO since 2015, and Laurence Madsen, head of IR and ESG. Thanks, Mark, for being here today. No, it's really good to be here. Thanks, Bryan. For the audience, before we kick in, if you want to have any questions inserted into the conversations, you can do that, insert them into that MeetMax platform, or you can email me directly, and with that, Mark, we're going to get into it. Naturally, we're going to have to talk about GenAI. We'll talk about Endava Flow, your partnership efforts, but before we begin there, I just want to start at a high level talking about growth and business performance, so the industry's been choppy, obviously, to say the least. Endava's experienced that. Before we get into kind of the current demand sentiment, if you kind of just step back first and think about the last two years, just talk about the root causes that you believe have caused the volatility. There's external factors, but also internal changes throughout the organization from M&A and large deal targeting, et cetera. I'm curious if you feel that you've gotten to a point here where those changes have been fully digested and you have a better handle moving forward. Let's start kind of at that high level first. Yeah, it certainly has been a choppy two, three years. I think, in common with most service providers, I mean, I think the macro has been challenging. I think that's true for everybody. I think we've been impacted by that, especially with the higher interest rate environment, that the amount of spend on discretionary, as we call it, IT projects, acceleration, which is what Endava tends to do, has been sort of suppressed through that. But I think the biggest sort of shift has actually been the impact of AI. I know we'll probably go on to it in more detail, but I think that has caused quite a pause, certainly with spend with providers. Our sort of sense over at least the sort of two years or so that corporates have been experimenting with AI, they've maybe diverted some investment from what they would have spent externally on innovation to internal work, where they've been experimenting with it, for want of a better word, proof of concepts, et cetera, and investing in licenses and people. We think we're probably getting to the end of that sort of phase where the need to show some payback, certainly from the C-suite, there's a growing sort of frustration in terms of, is this investment actually going to deliver work for us? Now, we have been leaning into that in terms of AI native and developing our own AI methodology. So we believe that we're going to be well positioned for that swing when it comes. So I think those are the sort of external factors. In terms of us internally, I sort of hinted at the first one, which was focusing on AI. So that's been an investment for us in building out Endava Flow, which has impacted profitability and margins. And I guess also, you know, we did our largest acquisition ever with GalaxE about two and a bit years ago. It's been a good acquisition. It's been integrated well, but it has focused management attention. So it's had a part to play in our sort of story. And I think also, one of the things in terms of, I suppose, criticism of the company has been our visibility and our guidance. Up until recently, we have tried to forecast the timing of large deals, which we think is a key driver of growth for us, particularly through the use of AI. And we have found that very difficult, alongside others who have also found the time taken for deals to complete being extended. So there's a whole host of factors behind it, some particular to us and some more general in nature. Okay, all right, that's clear so now, if we pivot to the actual demand backdrop, and specifically, if you can kind of talk about the progression you saw over the course of fiscal 2025. Yeah, so I think we've suffered from lack of visibility, to put it sort of bluntly. There are a lot of, certainly the larger deals for us, which is anything over sort of a GBP 10 million total contract value. The progress of those opportunities, those deals has been pretty slow. It's not been due to competitive processes in terms of changing sort of win rates. It's basically due to client decision-making being held up for sort of various reasons, of which partly, I think, AI is to do with it. So that's caused us to do a big pullback during the earlier part of the year and struggle with guidance. We now have the methodology. We don't factor those large deals in until they are won. So we won three last quarter, which are now part of the guide. But I think more generally, on the larger deals, we think it's firming. We've been here before, but we think actually the demand and the pace at which things are moving has a more predictable cadence to it and engagement from sort of management teams. So it does augur well for calendar year 2026, but the word of caution, we've been here before. So we need to be cautious on that. And it's certainly not reflected in the guide, putting deals that are in the pipeline into that forward guidance figure. So I think it feels like things are firming for us. I think certainly on the big deal front, I think on the non-big deal, it's a little bit puts and takes. We were at the bottom end of the guide or below the guide last quarter. And part of that was the non-big deal cadence and pipeline conversion. So there's still slippage going on, but it feels like it's stabilizing and getting firmer. And I think the bigger arc is work that corporates have been potentially insourcing in terms of the work that Endava does. Is it moving now to the service providers? I think it feels as though tentatively that is happening. Okay, okay. All right, so firming, but some guarded optimism there. We'll see how that plays out. When you think about the conversations, is there varying performance and tone coming from, I guess, verticals or geos? Kind of think about outperformers versus underperformers for us. Not really. I mean, interesting. I mean, we've been very heavy in financial services and particularly sort of payments. So again, in the bigger deal space, there are a lot more conversations in the payment space compared to 12 months ago. Now, they still haven't landed, but they feel a lot firmer, and just to remind everybody, we've seen like a 50% reduction in revenue from the payment space over the last two, three years. It's been a big headwind for us, and again, that sort of spills over into the banking and capital market space, so it feels a little bit firmer there, but we're also seeing, certainly on our larger deals, opportunities from other sectors as well, where clients are now starting to think about taking meaningful sort of steps forward with their technology investments. Okay. So your fiscal year wraps the calendar year, right? So you said June fiscal. So you've guided with a component of 2026 calendar in there already, so six months of that. Based on how you built that up, can you talk about some of the conversations you've had with clients and how they may be thinking about next year's IT spending plans, maybe relative to 2025? What are they thinking in calendar 2026? I think it's firming. So things that might have been conversations about clients doing things at this stage in the calendar year, I think were just that previously. It feels like there is a lot more activity to actually, which is the primer for real spend. I think that is happening more. There haven't been real indications. I mean, when we tend to engage with clients, certainly the bigger end, they tend to be more of a, what's the right word for it, a cost shift. So they will move wholesale certain things to us, looking for efficiencies. Now, those conversations have got a lot of legs at the moment. So I think the corporates that certainly we deal with are seeking to make those sort of changes during the course of FY26. There's more a sense of urgency to it. We still get things taking a little bit longer than we anticipate. You think they're going to make a decision, and then you have yet another sort of meeting to clarify things and the way forward. But as I said earlier, they're moving to a stricter sort of timetable rather than things going into a hiatus for an expected period of time. So there's sort of, I'd call it hardening of demand. That's how I'd refer to it. Okay. Hopefully, the geopolitical surprises are minimized to allow for that to thaw. Yes, yes. I mean, certainly things like tariffs didn't help during the year. We certainly saw a slowdown on our automotive sort of sector, but that seems to have blown away. I think people are dealing with it as part of the new normal in their decision-making sort of processes. It certainly feels that way. So I wouldn't say there's confidence, but there's this willingness to get on with it, irrespective of external events. Yeah, I've heard that theme volatility is the new normal. Just get on with it. Yes. Okay. All right. Larger deal. So you mentioned larger deal pursuits here. Some of these very big, actually. And you've won some recently, the GBP 100 million in an AI, I guess, driven transformation you talked about last quarter. But I guess when you step back, think about the pursuit of these large deals, what drove that? Was that a function of just that's how the market evolved versus a conscious decision by Endava? And how does it differ from the heritage pursuits in the product engineering base you've had? I mean, generally, we're a certain size now where it's big deals basically addressable and putting those sort of frameworks in. We want them because of forward visibility, witness the guide trials that we've had, but to a certain extent, it's also been about anticipating what AI productivity will do, and we'll probably touch on Endava Flow. But if you think about what our business model at the moment, we are about 75% is time and material, and the remaining 25% is fixed price. Some of it is outcome-based, but we deliver at the moment fully through time and material. Now, there's been discussion about the productivity that has arisen from using tools like GitHub, Copilot, et cetera, and these lend incremental improvements in velocity, 15%-30% sort of productivity, and under a sort of time and material model, the client basically sort of captures that velocity and productivity. Now, with Endava Flow, and we'll go, hopefully, touch on it in more detail, where you're using agents to deliver work, the velocity at which you can deliver is 200%, 300%, 400%, et cetera. So if you were contracting under a time and material basis, that would cannibalize your revenue very, very sort of rapidly, and hence why we've talked about our outcome-based large deals, and that at the moment takes the form, say, with a Paysafe, we were just talking about the GBP 100 million, is where we have a take-on contract. It's not just about that because it's about minimum spend, but that's the initial sort of form of it, where you're paying back the services to the client at a fixed price or sometimes a reduced price over a five-year period. It's part of the work or the framework. Those are the initial shape of those contracts. But to really secure the profitability, you need to deploy Endava Flow. You need a five-year period where you're taking on board initially the people. You can optimize it through T&M, but then you need to move to something that is more radical. And as you're doing that, you need a fixed price outcome-based sort of component. You're only really going to get those on these larger deals. We were anticipating this as part of the shift to AI, while we and others were also thinking out, well, what is our distinctive delivery model going to be in this new era? It's not just about efficiency because we've become an AI-native business. We know the issues it's going to sort of throw up. So we can engage with clients and take them through that evolving picture, rapidly changing IT environment. Okay, okay. That's clear. Last one on kind of the demand and the guidance before we shift over to AI, so just as you built the forecast for fiscal 2026, can you kind of walk through some of those underlying considerations for the year? And we always get that question of how much is contracted, committed, things like that. Any flavor you can just share on how you built this up and thought about the progression of the year? Yeah. So as I said, with the guide, we don't include the big deals until they're signed. So the three deals we announced on the call are in our guide for the second half, which is when they basically sort of step up. The approach is the same. We do everything sort of bottom up. We look at the contracting and committing pipeline, apply some skepticism, and we're applying a double amount of skepticism to pipeline conversion in terms of the guide. But the sort of simple math, the way I think about our guide at the moment, we did 178 in Q1, and we had an exceptional credit to the tune of about GBP 2 million. So Q2 would be about 180. That's a figure within the range that we're giving at the moment. So you have a first half that's about 360. So even if you were flat, you're going to be at a figure of GBP 715-GBP 720. And then you layer on our big deals, which is about GBP 15 million. So you're getting towards the GBP 735, which is the bottom of the range, with a little bit of growth. To get to the GBP 750, you need more growth, which again comes down to assumptions about pipeline conversion. Now, when you put that math to one side and look at how we've built it up in terms of what's the proportion of pipeline and contracted and committed for that full year figure, contracted and committed at the low end is about 81%. It's about 79% at the low end. If you compare that with historically how we've done that, it was usually around a 75%-73% sort of range. So we're trying to apply a higher level of skepticism around the pipeline conversion. We can't eradicate pipeline entirely because it just would not be how things play. And so it's this balance that we've got to strike. So I think that sort of logic gives a sense that there is the step up, but it is mainly driven through large deals. And then to get to the top end of the guide, we need higher levels of pipeline conversion in the non-big deals than we're factoring at the bottom end. Okay. All right. That's clear. Let's shift over to GenAI now. So first, maybe at a high level, you just talk about how the conversations and the actual demand shifts have occurred in specifically, again, GenAI and Agentic over the last year. And I'm just kind of thinking about the pace of actual scaling outside of just the science projects and the proof of concept kind of work. So I think it's taken us and others a long time to work out how you will deliver in this environment. Because I think if we go back sort of 12 months ago, we were talking about agents and how to use them. We'd even developed our own large language model using existing sort of technology. But the hyperscalers are investing significantly more than we could ever do it, and it's evolving so quickly. So again, the approach for us is to use best-in-class tools to be agnostic. We're a sort of services business. But actually putting those tools together has taken some time. And part of it, I think, is it's moved very rapidly, AI, and it continues to evolve very sort of quickly. But there's an element of, it's not probably the right word, but stability coming about. People know the broad shape of what it will look like and what it will do, and so we've been thinking, how do we put that delivery methodology alongside our existing distributed agile, the TEAM time and material model, which we call TEAM internally, so we've got Endava Flow. It is a methodology. It's like a workflow. It starts with signals where the models absorb public information, artifacts from clients. They then produce analysis that comes up with potential solutions, but then can flow very quickly. We're going to use flow quite a lot here into passing over to agents to produce the product. We call that govern because the humans are in the loop, and then it evolves and it moves very quickly, and I think because of the speed at which it does it, the old agile framework isn't applicable anymore. And I think this is quite different, I think, to what others have talked about. They talked about capacity or platforms. This is a sort of extension of our ideation to production sort of philosophy, where we're with the client, we're solutioning, and then we build sort of seamlessly. This just happens a lot more quickly. So the formulation of the methodology sort of Endava Flow, we have discussed it with clients. We have a lot of interest. At the moment, it's just out of the lab, I would sort of say. We have used it on internal work. We build accelerators, which are automations for what we do. So we've used it to do that. We've used it in a sandbox fashion with some of our clients to show them what it can do. So it's gone from the PowerPoint to let's see it in action. It's also, to a certain extent, part of the sales process now. If you're able to demonstrate and have a distinctive sort of methodology and approach, it sparks client interest. So I think it's a delivery methodology. We do build and incorporate AI on what we build for clients. And as I said, over 70% of our people are using things like GitHub, Copilot. But Endava Flow is the next sort of step in terms of the productivity. And just to sort of round it off before we go a little bit deeper, it's been part, is probably the wrong word, but let's call it that with clients in a sandbox sort of fashion. It is also in a stage where it's being rolled out across Endava's delivery capability and also looking at how it sits alongside TEAM and TEAM that are distributed agile sort of process, so it's in those early days, but what I can see it doing, we were talking about time and material and fixed price revenues, it could move to 50/50, 60/40. I'm not sure about the timescales for that, but the cost of delivering, at the moment, 95% of our cost of sales is people. There's still going to be humans, let's call it in the loop, working with the agents, but that proportion could be 60, 40, 70, 30, or whatever, and then it's the cost of actually running the agents. That's part of cost of sales, so it's quite a big shift. I think it will be margin enhancing, but it's in sort of rollout phase this year. But it's not in our figures in terms of that productivity uplift. This is something that will start to come through in FY27, our fiscal 2027 and fiscal 2028. Got it. Yeah. It's hard to peg the adoption pace, right, for you to forecast upon that. That makes sense. Yeah, and it is quite a radical change. I mean, we will still have project accounting, but the costs it collects, there'll still be time because there are people, but we're collecting other costs in terms of how the outcomes for clients are delivered. How prevalent across your delivery base is flow being leveraged so far? Very little. It's basically in the sandbox. We build it into some of the deals that we're bidding for. So it's in future sort of years that it comes in at, call it, full strength, but we're in the sort of rollout phase at the moment. Okay. There are elements. We've talked about the sort of investment we've been making this year, which is largely around software, but it is also about bringing people in, so there's been a particular sort of investment in, let's call them young graduates who've only known one way to work with technology, which is through AI, but can they deliver in an enterprise environment? Can they use Endava Flow? No, they have to learn how to do that. Okay. I guess it's hard to say what the financial impact will be. You're just having some test cases so far as I think about the transition from how you used to do things to this new model, revenue market. Yeah. So and then, how does it? Or how it's in. Yeah, and how does it sit alongside? I mean, we're not going to sort of partition the workforce into you lot do Endava Flow and your TEAM and TEAS. It has to be more fluid and seamless than that because you'll, and again, it's like the clients. It's like you won't, time and material isn't going to die overnight. It's going to be around for quite a long time, and so you need both methodologies there, and you deploy the methodology and the capability in the right situation, and going back to why big deals, we think that is, at the moment, that is a right place at which you deploy it because you've got an outcome. It's fixed price, and how we deliver that, bluntly, would be up to Endava. I mean, obviously, the client's going to want to be comfortable around quality and security, etc. But it's how we sort of deliver, and that is what is going to drive value for our clients and for Endava. Got it. Okay. Okay. Makes sense. All right. Why don't we take it back up to kind of the aggregate company level? Deal economics, pricing. You consider some engagements moving to these agentic-led models. You still consider overall demand somewhat constrained, albeit maybe some improvement on the margin here. As you kind of wrap it all together, how would you characterize the pricing backdrop right now? I think for us, I mean, you're talking about the two-year journey. We're still a time and material shop, basically. So it is all about the day rates that we can secure and the average rate per workday. That has been very stable for us. We haven't lowered it, and so they criticize us for that, lowered it to secure work. There is sometimes some flexibility at the edges, but keeping pricing discipline has been key for us. We may have caused margins to suffer somewhat, but we are going through this pivot at the moment. So it feels as though it is, again, it's another word, stable from a pricing position. It's still competitive. I still think there's less talk about macro headwinds, but the corporates are still feeling operating in a difficult trading environment. So it's still quite keen. But for us, it's maintaining that sort of pricing discipline as we shift the model. And then you've talked a lot about investments that you're making across the organization here. When we kind of step back and think about margin, you've got these ongoing investments in tech and talent, go to market. As you look ahead, maybe, I guess, in fiscal 2026 and beyond, what are the main levers for you as you think about recovering margin? I think our guide, certainly by Q4, implies some improvement in certainly the gross margin, not by Endava standards. I think we're about 30%. We'd hope to get up to sort of 35% and beyond from a gross margin sort of perspective, and part of that transition will be this shift from fixed price to T&M. I'm not in that sort of outlook implying there'll be an improved market in terms of rate rises, but I think there is, we're getting to a tipping point where I think work starts to move towards the digital sort of players, and I think there's just this feeling of a shift and this hardening in demand, and that gives some pricing power back. The big sort of uptick, I think, in profitability will come through Endava Flow, I think, as the revenues grow as well as corporates adopt it on a more enterprise basis, AI, rather than, let's call it, the experimentation phase, I think, that we're at the moment. That creates a big sort of tailwind that we should be able to secure. The adjusted PBT margin, which the operating profit level, we have great operational leverage there where the revenue, the margin over a fixed cost base brings outsized profitability improvement. Based on what you see and based on where you were in the past, is a return to that 40% adjusted gross margin, 20-ish% adjusted PBT margin, is that still in play, or are there any structurally different considerations that wouldn't allow for that? I think it's still in play. I would temper it back. I would say like 35% gross margin and 15% because, as you can tell, it's early days. I can't sort of describe to you how we will charge and agree these outcome-based sort of deals. But I cannot see if we do this right, execute on the rollout of Endava Flow and deliver the benefits to client-wide, it shouldn't recover to those levels. Okay. All right. Very good. Well, we've reached our time here, Mark. I appreciate the conversation. Thank you for the opportunity. Thank you all. Happy holidays. You. Good to talk to you. Bye.