To my left is Rachel Laycock, who's the Chief Technology Officer, and then we have Erin Cummins, who's the Chief Financial Officer, who joined us from London today. So we very much appreciate you taking the time to do that. You know, Erin, I want to start with you. Just at a high level, the demand environment that you're seeing, you just came off a pretty good quarter, but it's very interesting to see kind of where you guys sit in the value chain. So if we could start at a high level, and then we'll work our way down. Definitely. So if I perhaps talk about a little bit about Thoughtworks. Sure. So Thoughtworks has been around for 30 years now. We're about 11,000 people in 18 countries, and if I would just say quickly what we do from a technology perspective, we go all the way from strategy to design to engineering, and now even operational support for digital technologies. So the takeaway from that is that we span the life cycle of technology, and we we I really would say we're a very global business, a very diversified business. So from the demand environment, you know, the last year, we've seen some challenges, some headwinds, and we just, of course, reported our results. And what we're pleased to be seeing now, which is a change from Q2, I would say stability. Mm. So, you know, we've been in a place where we've been managing through various headwinds, and I think part of that is the global nature of our business. We are very diversified. About a third of our business is in APAC, and it really means the cycle that Thoughtworks is going through is a little bit different than some of the IT services. So what we're seeing now is relative stability, particularly compared to where we were in Q2. We're seeing less client churn. We're seeing less project churn. The back half of Q2 had a number of surprises. Mm-hmm. And we're not seeing surprises like that. So I would say on the back of what had been a difficult quarter, things are more stable. Obviously, this is the time of year where everybody is talking about budgets and looking forward into the next year. And just relating back to last year at this time versus this year, last year, it felt that most of our clients, many of our clients, there was a lot of discussions around budget cuts, how do we reduce team size, getting an overarching view from the CFO that costs have to come down by X percentage points. And so that was very much prevalent in last year's conversations. This year, it's more positive in that it's not discussions around cuts. It's been more stable from that sense, so we'll take that as, you know, a first good sign. But that said, our clients do still remain cautious. You know, the macro continues to be something that our clients are talking about. The sales cycles are still elongated sales cycles. The deal sizes are still smaller in nature. We are still seeing incremental ramp-ups. It is still that environment, but the environment is stabilized. And then there's more early-stage positive discussions happening. Yeah. Not quite ready to call the bottom yet, but it does sound like stability from the quarter, and what we're hearing today is starting to kind of creep into the system a little bit. You know, you did win 34 new clients- Mm-hmm. -in the quarter, which I thought was a pretty good number, considering, the environment and the backdrop. And I think one of the things you called out was that 51% or so was kind of from this, you know, sales motion that you guys have been working on, increasingly. So when we look at those deals, and you kind of touched on it a second there, but, like, what's inside of those? Are they... It does sound like they're smaller, maybe not as big and transformational as they used to be, but just the scope of work and the type of work, how has that changed specifically? I think in terms of the overall scope of work, it's a very generalized statement, but I think it's fair to say that, versus where we were two years ago, clients are much smaller. Their appetite is smaller for the big programs of work starting. Mm-hmm. I think consistently what we're seeing is our clients wanting to see proof points around ROI, and so that's definitely thematic, having more layers of approval associated with the overall—s o if we look at some of the work that we do with our clients, it is inevitably long term in nature, where the client discussion will be, "Hey, we've got something. This is the impact we're looking for on the business. We think this could go for two or three years, but right now, we're just gonna start with this piece. We want a smaller team. We need to demonstrate ROI. We need to show this to the board and make everybody comfortable with the value that we're able to drive before we sign up for something different." And so that is pretty consistent across the portfolio, and it definitely is with the new clients, and so we're incredibly excited about the new client wins that we have, just some amazing logos. The conversations are very long-term. There's a lot of potential and excitement about what we'll be able to do with them for a longer-term horizon, but right now, people are still in that, "Let's start small, let's make the proof point, let's build trust across the organization, and then we'll go from there. Yeah. I feel like, honestly, that was a competitive advantage for you guys even before this. You always kind of prided yourself on being able to deliver that ROI in kind of smaller sound bites, and then that seemed to parlay into bigger digital transformation, which I'll lead to you right now, Rachel. So digital transformation means a lot of things to a lot of companies, and we cover a bunch of different services companies, and depending upon who's sitting up here they'll give you kind of a different answer. But where you sit, again, in kind of the digital engineering space, how do you define digital transformation for your clients? Yeah, so one of the unique differentiators of Thoughtworks is that we are end-to-end. So when we think about digital transformation, we do think about top to bottom. So everything from the customer experience and product design, to the underlying, like, platform design, which we've been doing, you know, creating really differentiated platforms and digital platforms and engineering platforms for clients for years now. We also look at, you know, the operating model making sure that it's frictionless. So a lot of more, I guess, older operating models can be quite slow. You'd be surprised at how many organizations still have a lot of Waterfall built into their systems, and checks, and gates, and whatnot, so changing that is really important. Then, of course, you've got the underlying data. And I would be remiss not to talk about AI at this point, because it's the topic that's very hot, but it's something that we've been helping clients with in terms of gaining insights from as they're modernizing their data and introducing data products and Data Mesh is something that we've been doing with clients for years. So we look across the whole spectrum, and then, of course, that we underpin that with our, like, long history of- Mm - you know, engineering excellence and, and getting to value. We, we believe that all of those different aspects, and regardless of where we start, to Erin's point, within an organization, we try and look at the whole picture, and create a roadmap for a client of like: "Okay, let's start here, let's get to that value, and then we'll look at the other aspects that we have to, we have to change along the way. Yeah. But we do try and look at it holistically, and make sure that, you know, we're getting clients to value along the way, because these are long journeys. They are years, for most organizations, and if you don't demonstrate value along the way, people get tired of it. They start to ask: "What's the ROI? What's the benefit? Mm-hmm. Things get stalled. Often, in the middle of a transformation or the middle of a modernization of a system is, like, the worst place you could be, because that's when you end up with, like, two of the same system. That's why you have so many organizations where they'll have... not just through acquisitions, but they'll have, like, three or four of the same system because they've tried to rebuild something, and then, you know, the business has got kind of exhausted by not seeing that ROI. We just think that is so important, and that's why we focus on that, but look at it holistically. Yeah. So that kind of leads to my next question, which is, you know, balancing this experimentation that maybe clients want you to go out and explore, but also delivering that tangible outcome. So you lead with the ROI, or is that something that kind of comes later? And that could be a question to either one of you. We try and do it as quickly as possible. Yeah. It's interesting. It kind of depends on the organization, where they are in their maturity ' cause not all technology parts of the organization do have a heavy ROI, like, value focus. Mm-hmm. So sometimes that's part of the transformation. But we pride ourselves on experimenting and then harvesting that and scaling out, right? Like, we were pioneers in microservices and DevOps. We invented Data Mesh and the whole data as product thing, and that's literally just experimenting and innovating on clients and being like, "The current approaches, they're not solving this client's problem, so let's look at another way to solve that, and get the benefit for the client." So that's why we always kind of say things like... We don't like the term best practice because it's like, is it the right practice? Right. Is it the right method? Is it the right thing for this client? And if it's not, can we find a better way? And that's just built into our culture to try and get to value. Yeah. If you wanna add to that. I would just touch on that, to add to Rachel's point, to the ROI. We always would like to get to a place where we can clearly say, "Here's the financial ROI. Mm. But some clients, we can get there because they are further along the maturity chain. But for other clients, it's, "Hey, we've got this program, we've got this, this move to the cloud that has to be done, this application that has to be modernized." So it's very specific. The outcomes are known, and clear, and so there's a value in that sense. And then, for others, we're along the journey where, we're trying to get to a place where we can actually put a dollar on certain things, and, and that's something we work with our clients on. I feel that that's probably— you know, it's less than half of the clients that actually- Yeah -you know, we're able to do that with, but it's the place that we continue to focus on and see more and more. Okay. And can you spend a minute and just talk about how you build the kind of backlog funnel or revenue funnel, however you wanna describe it? You are very close with your clients. You have discussions with them regularly. That's a unique position that brings a lot of insight. So how do you, how do you think about building that as you sit here today and you're talking about budgets in 2024, and the visibility that you think you're able to garner from those types of discussions? Sure. So, you touched on a key point, Dan. You know, we're having the right conversations with our clients because we're engaged in the strategy discussions. So that gives us a huge amount of long-term visibility, because a lot of the time, we are helping the client build that technology roadmap. And so not only do we have huge value to add there, but it is absolutely a win-win because it helps us understand what's in front and make sure that we're working on the most important things with the clients. So that creates enormous visibility that doesn't necessarily always translate immediately into- Yeah -revenues or contracts. But on the whole, if we look at the structure of our revenues, over 90% of our revenues come from existing clients. Yeah. If we look at our portfolio, we have a very well-diversified portfolio with not too much concentration, and the top ten— we have an average tenure of 9 years, top 10 clients. But in reality, if you look at our top 50 clients, it's a similarly long tenure. So the clients get to know us, we build a trusting relationship, they understand the value, and we go from there. Now, in terms of building the funnel and the process that we go through, every, you know, every client has a little bit of a different flavor, but typically, you know, there's existing programs of work that we'll continue pushing, executing. And I'll give one of our larger clients an example, where we started a program, in APAC many, many years back. We built particular technology, furthered that technology in Europe, and we're helping them roll this out over, many different geographies for their business. So really it's something where they've said: "Yep, the value's here. This is really good." It's in automotive, and so we're rolling that out. And so that's a case where there's years of visibility of a longer-term project. With some of our clients, particularly where they're the mature scale-ups, Thoughtworks is an embedded part of their technology. Mm-hmm. We might say, "Okay, well, we know that we're going to have 150 people from Thoughtworks working for the whole of 2024. We don't know exactly what they're doing, but we'll figure it out as we go along, but we know we want them as part of the team. Yep. And then with other clients, it's especially ones that are newer, you know, we've had the initial win, and we're building the strategy, and really what we want to get to is the delivery. And so we'll have visibility over that. The delivery piece is inevitably going to be the larger piece of it. But that will be a key part to our pipeline as well, and moving that forward. Okay. Okay. Well, we held off long enough on GenAI. Yeah. We did it. Fifteen minutes. Did it. I mean, I was working really hard to do that, by the way. What I wanted to start with you, Rachel, is the question that... Are there, like, technology prerequisites that are required for organizations to just start down that journey and be able to extract the real value, you know, out of GenAI? And then how do you approach that? So, I'm going to give you a very consultative answer, which is the "it depends" answer, but I will give more color to that. Okay. So if you want to use, like, a GenAI large language model from a hyperscaler or open source, then the prerequisites are not as high. But if you... And then there's a spectrum. If you want to build out your own large language model, then obviously there's the cost of that, but there's also, what state is your data in? Like, have you already gone down the- Yeah -h ave you moved it to the cloud? Have you gone down the data products and Data Mesh approach? Are you able to access the data to train the data? And then there's everything in between, which is like leveraging a large language model, and you can deploy, you know, models on-prem. But it starts to. There's a question of, like, how much accuracy do you need? How much can you afford to spend on it? Yeah. What's the value of it for you? So for example, one area that we're working in is in pharma, where, as you know, drug discovery is an extremely expensive process. So for them to build their own large language model, train their own data set, it's basically a line item on an already extremely expensive process. But for other organizations, it's prohibitively expensive. Yeah. So they're going to look at either leveraging the language models of the hyperscalers or, you know, leveraging it to create their, train their own data set. So, what we're offering, which I think is unique, and our clients are telling us this, is that because we're very agnostic, we pride ourselves on being such, we, we have relationships with all the hyperscalers, but we're not tied to any specific one. And so for some clients, you know, who are saying, like: "Some of this stuff is on-prem, and we need to keep it on-prem. Mm-hmm. We look at, like, open-source models that they can leverage and in many different ways. But it, the prerequisite depends on what you're trying to do with it, right? If you just want to start using, you know, GitHub Copilot and some of the off-the-shelf tools that are available, the prerequisite is low, right? Yeah. You just start using them, you see what you can do with them. But once you're actually wanting to start building your own products, build your own large language models, then the prerequisites get a lot higher. Yeah. You bring up a really good point. Cost is definitely something we hear all the time about how expensive, you know, these types of models are. But you know, the thing that you brought up, which made me kind of pique a question, is really, like, when companies have huge budgets like pharma, then maybe this is slotted in there as a line item, and therefore there's a lot more internal cost that they can absorb. So as you think about having discussions with certain other verticals, you know, where do you think the bigger outsourcing opportunities lie? As in, would a financial institution vertical be one? Would industrial and production... Like, I'm just curious, as you're thinking around that whole spectrum, where that might fall out. We are getting a lot of interest from financial services as well. Yeah, and I understand. We're with that. Yeah. Yeah. I would probably say it is mostly in the life sciences, the pharma, like, space, and also, financial services, where we're looking at organizations, or organizations are exploring building their own large language models. Yeah. But I think for other industries, it's much more, they're immediately seeing that it may not be worth the ROI. Yeah. And so for example, one client that we've been working with, and this is one of the services that clients are really excited to work with us on, which is there's two. One is, like, GenAI for product discovery, so figuring out what products you'd actually get the benefit from leveraging GenAI on. Mm-hmm. And then the other is, GenAI for software delivery, because of our, you know, history and expertise. Like, we're at the forefront- Yeah -of basically redefining the software delivery process. So for the former, like, we have a—a nd I've recently did an exec roundtable with about 10 of our clients, and it and also a very specific client, looking at 270 different use cases. Wow! Most of them are in that analysis phase of, like, trying to determine what are the use cases, and then where can they leverage things that are off the shelf, that they can leverage from their existing partners, and where would they, you know, have to invest themselves, or where could, you know, on all the spectrum in between. Mm-hmm. And what I'm seeing is every industry, every client is looking at how to apply this, from both what products and services that they offer, as well as how does it impact how they build their own software internally. Both. They're looking at both sides of that. But I think, like, it remains to be seen where the real value lies. Like, I think I predict there'll be a lot of, I guess, duds when it comes to products that get built over the next few years, where it isn't worth the cost. Because- Yeah -as we know from just building products in general, like the complexities of choosing the right product, and did you overshoot from a functionality perspective? Do people really want that? Is it really useful? That is more of an art than a science, and it's a hard problem to solve, and many people have been writing many books about it for many years. Yeah. That's part of the reason that you bring in consultants to try and help you, you know, figure out what's the right products to build. That doesn't change. This just adds a layer of complexity. But if you wanna go, as I say, really deep into building your own large language model, training your own data, it's very expensive, so it has to be worth it. Yeah. So a lot more research needs to be done. One of the things that we've been talking about a lot is, anybody can build a proof of concept, and if... and in this case, if you build a proof of concept in the small, it, you know, that's fine, it works. But once you start thinking about, well, what about if, you know, 1 million customers are using this? Mm. Could we scale it? What does that mean for us? It's a totally different conversation. So I think this is something that, again, I think we're really in a good spot to help clients with because of our focus on value. So we're almost abandoning proof of concept and talking about proof of value, right? Like, okay, you've got an idea for a product, how can we get it into production and show that it's useful in some small case before we think about whether we scale it out? So those same concepts still are really, really important. That's interesting. Does that concept that you just introduced suggest that maybe the ability to get back to revenue recognition, and I guess maybe this might be a question for Erin, could be a little bit more elongated as these projects start to be absorbed into the financial model? I think that in the sense of so looking at our revenues, about 75% are T&M, 75-80% are T&M. Yeah. So, what we've been talking about and experimenting with, well, what is the right commercial model for, this, more advanced AI or GenAI work? And we are experimenting. It's a small piece to the revenue, but we are experimenting. So firstly, you know, whenever the strategy, the discovery, the assessments, you know, they do tend to be smaller upfront- Mm-hmm. -and so the revenue stream Yeah - it doesn't build until we get to that bigger phase of delivery. Mm. But then also, maybe what I would say, too, there's a stops and starts that can make it a little bit tricky from a revenue perspective, because- Interesting - you know, when people are trying to determine what path they're going to take, but they're in that experimentation phase, there is the need to say: "Okay, well, let's try this for a minute. We need to stop, think about it. We're gonna try this." And so those stops and starts are not always the most helpful to the basic T&M consulting model, if that makes sense. Totally. So that's a challenge that we're working through, and we're looking at different operational models so that we can manage that in different commercial models. The other piece that we're thinking about and experimenting with is, well, how do you define this value? Especially if we get to a place with the AI capabilities, where it is different if the unknown of what we're trying to get to is more unknown, I suppose, than we're used to, then how do we think about that from a commercial perspective? So a long way to say, right now, the impact on our revenue recognition isn't significant. It's not been a problem. But as the AI part of our portfolio, we do see the need to make sure that we're evolving that and thinking about it in a different way, because we do think it's gonna be a different model. Yep. Yeah. Yeah. We're modeling out, like, the team structures for, for example, Copilot, and the fact that it can write the CRUD operations, the create, read, update, delete types of things that, you know, is just—it's very time consuming, but it doesn't add a lot of value. Right. It can do that. But the complex part of systems design and design, and these tools can even help with requirements gathering, they can help with idea generation, they can help with—from a design and wireframing perspective... what that looks like, which is why it's, you know, this is a core part of our business. It's a number one priority from a technology strategy perspective, is like, looking at exactly what these tools can offer, building our own toolsets around this, and then, you know, looking—'cause, and also, I would say these tools are changing all the time. Like, we have early access to things like GitHub Copilot. They're introducing new features all the time that's gonna change the ways of working of development teams. Yeah. So this is a fast-moving space that we're right in front of, and constantly assessing, like, what's the team structure gonna look like? How do we price this? But it's gonna change quite a bit, I think, over the next 12-18 months- Yeah - before it settles, and we start to see what the future looks like. I mean, it's really amazing to think about the engineering aspect of that. I mean, that is a core part of your business, and so there's a piece of that that could become so much more efficient, but it's probably gonna have to pivot at some point to make that efficiency gain. And there might be some self-cannibalization that needs to occur, and so it's, yeah, it's amazing. All right, let me bring it back for a moment to the business. Not that AI is not fascinating, but, you know, the restructuring. We can talk about it. Yeah, I can tell, like, it's a passion. Restructuring project, you know, you guys put in place in the second quarter, maybe bring us up to speed there. I know you provided some numbers around last quarter, but it does sound like you're gaining a lot of traction there, and that's freeing up a lot of, I think, opportunities for you to continue to invest as we think about going into 2024. Absolutely. We, you know, I will say at the start, restructurings are not easy. I don't think anybody would call them easy, and ours isn't. We knew that it's going to be a challenge, but one that we feel is absolutely right for both the near term and the medium term, frankly. So the restructuring program we announced in early August, and our overall goal from a cost savings perspective is annualized savings of $75 million-$85 million. Mm. We at the close of Q3, so the end of September, you know, we had already locked in annualized savings of $68 million. Yeah. And so we feel really good about the progress. We have a commitment to working quickly because the change is disruptive for people, and, you know, it can cause challenges around change that are natural, and so we wanna move people through that cycle as quickly as possible and get to the other side. But Dan, what you mentioned, and ultimately, you know, where are we going with it? Firstly is client centricity, to be closer to the clients, to be more responsive, particularly on industry view, and we are absolutely already seeing that pay off. In fact, you can hear a bit embedded in Rachel's comments, it certainly helped inform our thinking about how do we apply AI and GenAI to our markets, and so that's been very helpful. What we've seen in Thoughtworks, and it's been part of the revenue headwind, is a lot of movement from onshore to offshore firm for our clients, and so that's been a big, a big piece that it's hard to see if you just look at the top line of the company. Mm. You have to, you know, disaggregate the situation to understand that more clearly. But we see, for example, in Q2 to Q3, we see the volumes of work that we're doing with our clients being very much the same, but, a lot of that work has moved offshore. And so fundamentally, we think that change, it's a change that's been happening for a long time. We don't think it's gonna go back, but it happened at an accelerated pace versus where our business was, and so we've made the adjustments for that in the restructuring program. And then finally, what we're seeing and, and what we're excited about, so we have this Digital Engineering Center, and it's really about instead of managing 18 different populations of people, and you can imagine, you know, what that's like, it's bringing that together at a global view, certainly a regional view. And, and so that will allow for efficiencies and allow us to, see a higher rate of utilization, so that we're not only... And where we're investing, clearly, we're investing in AI, but making sure that those investments are, in very specific places, rather than having, separate 18 different investments, if you will. And so we're seeing the utilization pay off, and we're also seeing the cohesiveness of the overall investment in our tech strategy- Yeah - come together, so. Okay. So in reconciling some of the commentary that came out of the fourth quarter, which I think was more around top line, which was stability from fourth quarter to first quarter, and then maybe ramping from there- Mm. How do we kind of interject the margin commentary based on what you just said around all these great things that are happening around the restructuring costs? So from a margin perspective, I guess, you know, this has definitely been a transition year and a year of a number of challenges. Again, the accelerated move to offshore has impacted margins. Yeah. As we look forward to next year, what I can say definitively is we do expect a higher margin next year. A lot of that to be coming from our improvements in utilization. That said, as much as I wish I could snap my fingers on the 1st of January, say, "It's all done," you know, that'll benefit the year, the truth is, it won't. It's going to take a couple of quarters to bake in all the changes and see the shifts that's happening. So if I consider the margin cadence or trajectory that I expect to be on, it's one of seeing increasing improvements to margins. Next, you know, then our company has performed over the last couple of years in the high teens, low twenties. I don't think that's what we're looking at for next year, but I do think that we will get back to that place in the medium term. Okay. Well, we're just about out of time, so I'm gonna leave it there before I start another discussion on AI, that'll take us over by a wide margin. But Rachel and Erin, thank you so much for being here. Really appreciate it, and I am very much looking forward to how 2024 kind of plays out for this company, so- Yeah. Thank you very much. Thanks, Dan. Thanks. You too. I'm-