Good to have Jeff Puritt, who's the President and Chief Executive Officer of TELUS International, joining us today. Yeah, good morning. Jeff, thank you so much for being here. Happy to be here. You know, what I thought I, as, as we're gonna probably do across the entirety of the day, is really start at a high level, talking about the demand environment that both you and just the environment that you're seeing is operating in. You know, you're fresh off of the third quarter, so it's really compressed here. But what I was hoping to start with is just your perspective on what's really happening across each one of the maybe three big areas of a focus for you today. So from a macro perspective, I wish I could say that we are enjoying a return to normalcy and the double digit, starting with a two or a three revenue growth profile that we were enjoying for years, up to and including our IPO in 2021. Yeah. Unfortunately, for us, the practical reality is we're still seeing a challenging macro dynamic at play here, with slow decision-making, elongated sales cycles, and a diminished demand environment in totality. There are pockets, areas that have some variability within them. So, for example, unique to our business, I think in part as a result of a undue client concentration profile that we've been talking about for quite some time now. Mm-hmm. As we've shared, both in Q3 and the quarter before Q2, one of our tech and game social media clients, where we're providing trust and safety content moderation support, principally out of Europe, significant diminution in demand, and unfortunately, we hoping to repurpose our skilled digital first responders to other opportunities, simply didn't manifest itself fast enough. And so we were slower off the mark than we should have or could have been, and as a consequence, Q2 was weighed down significantly. Fortunately, we were able to make meaningful progress on right-sizing, matching labor to demand in Q3, and so you saw that reflected in about a 370 bp improvement in margin yield quarter-over-quarter. Not exclusively off of the labor right-sizing, but meaningfully so, together with some other austerity programs, if you will. Mm-hmm. One other sort of negative in terms of demand from our BFSI sector, where, again, a single client in the U.S., they themselves having less success in launching new support in turn led to a diminishment in demand for our support. And again, you saw that in terms of a softness in that vertical for us in the quarter. Other than that, though, I mean, pretty steady as she goes in terms of growth. If you normalize out that social media decline for us, we had 13% revenue growth year-over-year for the third quarter, 17% in the tech and games vertical in particular. Yep. So we think there are pockets of opportunity, and most excitingly is our TI AI digital solutions portfolio, where we've seen 30% growth in our customer base. You know, 25% growth serving Google alone year-over-year for the third quarter, helping them, Microsoft, and a number of other Fortune 50 clients building their large language models in particular. Boy, does it feel like we were lucky / prescient making that Lionbridge AI acquisition back at the end of 2020. Yeah. But we think there's pockets of opportunity, going ahead and having made those investments, whether it was LAI, even more recently, WillowTree. Challenging year so far for sure. They, like we, similarly affected by these elongated sales cycle and a bit of, I don't know, paralysis in terms of decision-making and commitment to large, transformational projects. But I'm hopeful that three years from now, we will look back on the WillowTree deal, and it will have been as strategic and prescient, as the LAI one has turned out to be for us. Yeah. You know, when you think about, you know, building the funnel back up, so to speak, or keeping the funnel full, and the conversations you have with clients, which seem pretty, you know, you're very close with them in terms of their discussions about what they're thinking about. So what can you tell us about how, as you're building out that funnel, the business, the backlog, over the course of the next 12, 18, 24 months, as you're visioning it, where are those discussions these days, and how maybe are they different from... And I don't even know if it's a time period. It could just be whenever it felt normal to where we are today. So our funnel continues to be rather robust. We're still well above $2 billion- Yeah. ... of active opportunities, qualified opportunities. And there's been some variability in the size of the funnel over the last few years, you know, since we went public back in February of 2021. But, you know, kind of never lower than 1.9, not really above 2.8. Mm-hmm. So we're kind of in the same zip code. But what's different today, and has been this way now for going on, gosh, three, four quarters, is things are just not moving. Decision-making, getting to yes just seems to be taking considerably longer than it did back in the good old days. Yeah. And it's ironic, right? I mean, gosh, we went through a one-of-a-kind, hopefully, global pandemic and quarantine restrictions and, you know, so call it, I don't know, March 16th, March 19th, 2020, we all thought the world was coming to an end, and, and we had to figure out how to get, at the time, 60+ thousand workers home and productive. And we're fortunate to virtualize our entire operation quite quickly, and, and we came out of 2020 with very, very healthy double-digit revenue and EBITDA growth. So that we were able to overcome that seemed like, well, this should be, you know, a cakewalk, and yet this is different. This has been elongated, the uncertainty from a myriad of factors, sort of this perfect storm around interest rate inflation, and who knows if and when it stops, and supply chain constraints, and geopolitical uncertainty, whether it's Russia's invasion of Ukraine, now the Middle East. Yeah. It's like one thing after another, after another. And it's, I think, for us anyway, leading to a number of decisions not being taken on whether they're hoarding cash or otherwise. I mean, even in our own businesses, we're looking to constrain our spend, optimizing our supplier relationships, finding opportunities to do better or more with less. The very value prop we offer- Mm-hmm. ... to our clients is, you know, physician, heal thyself. Figure out how we, too, can be deploying AI-enabled capabilities inside TELUS International to reduce our cost to serve, create more headroom, so that now this pervasive desire for more or better for less that we're hearing and seeing from all of our customers is something we can be responsive to but still sustain, you know, what has been a strong legacy of double-digit EBITDA yield. Yeah. You know, we had an unfortunate bump in the road in Q2, but thankfully, we're already back to 2021 and change, and we'll exit the year, as Vanessa shared during our earnings call, sort of run rate around 23% again, and hopefully get back to normal and better profit yield going forward as we continue to leverage technology and automation in our solutions. But that's what's different in the funnel today. I mean, there used to be a sufficiently robust demand where folks wanted the high-end premium service and were willing to pay for it. Mm-hmm. Now, everybody wants that same high-end premium service, but nobody wants to pay the premium. Yeah. They want more for less. We're empathetic in our business, too. But, all kidding aside, so, you know, you mentioned the number three social media client, and that is definitely disproportionately, I think, kind of jaded in many ways, kind of what the whole picture looks like. But that relationship in particular still feels like it's going well. They're just... But they also seem like they may be further along in that process. So any insight into maybe if we're getting closer to the end of their more, I would say, cost rationalization or optimization viewpoint, I mean? So it is still a relationship of significant consequence for us- Yeah. ... obviously, given the magnitude of the relationship. I'm cautiously optimistic- Okay. ... that we have seen the bottom, and there's nothing but upside from here, but I need to be, you know, realistic here. We are not out of the woods yet, unfortunately. Okay. And so what you heard from, again, Vanessa, on the earnings call, was a degree of conservatism, in what we're expecting, or sort of with they, as well as all of our other clients, going into budget season, they and we, as we start to talk about what the demand profile is going to look like going into 2024. So I don't want to get ahead of myself. We'll provide formal guidance, in February when we- Okay. ... report our Q4 results. But we're hopeful that we are near the bottom of that trough, and there's more upside opportunity with they, and frankly, other clients just like them. Yeah. No, it feels that way. All right, I've held off long enough on AI. I tried, I tried my best, but here it is. Are we gonna do that- Here it is. ... that beer drinking game every time we say AI? Yeah, I mean- Take a shot? ... we might as well because it's gonna be very prevalent at this conference. So, so look- The beer drinking or the AI? Maybe both. The beer drinking typically takes place latter half of the show. But in this case, there's just a lot of discussion around content moderation, digital CX, just all of those things around what AI and generative AI is potentially going to do, but also the opportunities that it creates for you. So maybe help us kind of with some fact versus fiction, so to speak, from where you sit, because you're building the tools for clients, and they're obviously coming to you and asking for those exact solutions, so. Yeah, I mean, I guess for me, the headline is, as McKinsey has said, as Deloitte has said, I mean, there's few experts out there who have not suggested that generative AI in particular, not just machine learning AI- Yeah. ... is going to be among the most disruptive, events in history. I mean, just look at the adoption curve for generative AI tools to get to 100 million users- Yeah. ... in 60 days. Staggering, the pace of progression. Our business has not yet been as disrupted by GenAI as I anticipate that it will. But I am bullish on we being a net beneficiary of GenAI because, as I said a moment ago, we were lucky/ prescient enough to get in on the ground floor. The data annotation capability that we secured through that L AI acquisition that we made at the end of 2020, and the, you know, Crowdsource community of over 1 million folks that were helping, the hyperscalers in particular, but not exclusively, to build their machine learning algorithms. You know, that is an ecosystem in terms of the technology that underpins it, sort of the community management platform, which helps you to identify all of these workers, have their IP address, have their skills inventory, all automated, so that when new tasks come out from the end user customer, the hyperscalers of the world, you can quickly identify who they are, where they are, get them onto the platform, train them, validate them, have them submit their work, validate the work has been done properly, and pay them. Mm-hmm. That entire ecosystem has served us well now as the market has pivoted, not entirely, because there's still a lot of annotation going on. Mm-hmm. But there is so much more work now underpinning these large language models, which is more reinforcement learning from human feedback and/or supervised fine-tuning. And so those two are similar but slightly different activities that are used to help to ought to optimize these generative AI models. And we've been able to repurpose most of our CMP, the community management platform, and although the community that are actively engaged in these efforts is now circumscribed a little bit, because you want experts. So if the nature of the output for the generative AI model is a math question, well, you need accredited mathematicians who are participating as prompt engineers, building the questions and the proposed responses for the model, and then evaluating the helpfulness, the harmfulness, the accuracy of the response that comes from that generative AI tool before it's ready for prime time. What we've evolved to in building our own API experts platform is we can automatically identify the task that is being provided to us by the customer, help us optimize this activity, then automate identifying members of our community and our crowd that have the requisite accreditation to effectively perform the task that's required, and then the entire workflow is automated. It expedites that entire process and does it at a reasonably cost-effective level. We see that continuing for quite some time, because it's not just these massive large language models that the hyperscalers are working on, although they're the- Mm. ... the most disruptive and the ones that, whether it's ChatGPT, OpenAI, and others are producing. But we see that across the enterprise as really the future of generative AI. And the difference here is, when you put your data into one of those hyperscaler generative AI models, you just lost control of all of your confidential customer proprietary information. So that example from Samsung with those, software developers were, you know, inviting generative AI to help optimize their code, not realizing that once they did that, that code was now out in the public domain. Right. So being able to build large language models inside an enterprise private cloud environment, so their data never goes anywhere other than where they want it. But if you don't structure that data, if you don't make it meaningful, it's not accurate. You're only ever gonna get hallucinations, a polite word for garbage. Mm. We think because we've had so much experience now building these monster large language models for the hyperscalers, we have exactly the right capabilities to take that down just a little bit inside an environment, like for example, with TELUS- Yep. ... our parent company, where they've got voice recordings, and chat, and email, and service offerings, and subscriber plans, and annotating, structuring all of that data in a way that is accessible using these generative AI capabilities, so that, A, we can create an environment that helps improve the self-serve ecosystem. So TELUS can divert 60%+ of its contacts with its customers to a web channel or even a mobile channel, but it's all self-serve, and it requires no human intervention, so reduce the cost to serve and improve the client experience. Mm-hmm. While concurrently, if it is unsuccessful on a self-serve front, and they have to end up failing over to talk to a human being, that human being is instantaneously supported with this data-rich environment. You know exactly who you're talking to, what questions they asked the bot that were unsuccessfully answered, where they spent their time on the website, what products and services they're already procuring from TELUS, or perhaps from a competitor. And a generative AI assist bot will automatically populate the desktop of that agent with the next best recommended response to that customer to accelerate, expedite, and improve the entire experience. That only happens when you have access to that data. You can structure that data, create those interfaces, whether web or mobile, and then serve up all of these tools, these bots, whether it's language translation or assist bots, et cetera, in this private environment that ensures that data never leaves that ecosystem. Yeah. So one of the things we hear, and we hear this, like, internally at the bank, is just AI, the cost of building these things is, incredibly expensive. Yeah. One of the things that I'm hearing from you is the ability to port some of that cost burden away from the client, in addition to structuring this data in a way that kinda keeps it proprietary in a lot of ways. I mean, is that, is that part of the value proposition? Are clients even at that stage where they're thinking about porting costs to you, or are they still in test and learn? The vast majority is still dipping their toes in the water. Yeah. I think this is still the early days of the gold rush. You used the word cannibalization earlier. We've debated that word a lot- Yeah. ... inside our business. I think the reality is that we are going to be cannibalizing our legacy revenue models- Yeah. ... because the traditional bum in a seat is going the way of the dodo, in my view. It's not an if, it's a when. It's how fast and how pervasively. I think we will always have human beings in the loop for a whole bunch of reasons, including even in a traditional care ecosystem. But what those humans do is already moving up the food chain in terms of the complexity, sophistication, and value of the work they perform. And to my earlier example, they need to be near superhumans in terms of interoperating with, you know, it's as if they're a pilot on a spaceship with this 360-degree view, metaphorically and practically, of data. They still need to be articulate and empathetic and be able to hear what it is the customer is asking and respond appropriately, now armed with all of this information. But long gone is the simple, predictable, repeatable stuff. That's all RPA and automation, et cetera, et cetera. So for us, finding a financial business model that says, "You know what? We only ever used to generate dollars for people." And so time and materials was, you know, ubiquitous, and our pricing model for our new Fuel iX offering is a bit of a hybrid model that's both a monthly license fee as well as a usage consumption-based model. Mm-hmm. So it's fixed fee for professional services consulting to do the original... We have an eight-week jumpstart AI program that our WillowTree team are fronting as part of this Fuel iX offering, and they'll go in and talk to you about what your environment looks like, what we see in the world at large, and how generative AI, in particular, can probably help you take costs out of your service to your customers and improve the client experience concurrently. From there, again, on a fixed-fee basis, structure your data, provide sort of an actionable, monetized access to those insights, again-... not dissimilar from what I described a moment ago. Mm-hmm. Create these mobile and web interfaces that make it easier for customers to buy stuff from you and to get refunds, moves, adds, changes, et cetera, as well as to help your agent population support your customers. And then fourth is this really sort of a Rosetta Stone, metaphorically, this platform that allows for plug-and-play APIs. Our first offering includes 10 different feature functionalities, whether it's a translation bot, an assist bot, a chat bot, et cetera. And if you're already using Google Translate for translation, or you're already using a bot from someone else, you can simply plug it into our platform. If, while we're consulting with you, you heard about something else, well, we can procure that with you and include it in our price. By providing that hybrid pricing model, we're essentially cannibalizing the legacy revenue stream because this ecosystem will displace between 20%-40% of the human beings that used to be required in that ecosystem to manage the contact volume that you used to get. Mm-hmm. For us, the ambition is that while we may ultimately give up 24%-40% of the revenue run rate within an existing client base, A, the profit yield from this will compensate in part, and B, hopefully, we'll actually be cannibalizing the competition's relationship- Yep. ... much more quickly, more pervasively than our own. Well, you jumped into Fuel iX pretty quickly, and so and you kind of explained it, but I wanted to make sure we were clear. Like, this is the go-to product that you're coming to market with today, and so, like, what's the market motion there? And, are there anything, you have a dedicated sales team? Do you have, you know, a different strategy in terms of attacking it? 'Cause it's relatively new. I mean, you, you just introduced it, so I wanna make sure we're all clear in terms of- There's lots of- ... exactly is. ... lots of work to be done. Indeed, I think perhaps forgive me for jumping the gun- No, no, that's great! I'm pretty excited about it. That's great. We have a lot of work still yet to do. We have a number of pilot customers already on the platform that a few of which were quoted in the press release and that we're already working on, but dozens more than that. I wish I could sit here today and tell you that, you know, I can't keep Fuel iX on the shelf because it's flying off. Right. We're not at that point yet, but hopefully, next year I'll come back to this conference, if you'll have me, and I'll tell you exactly that. Admittedly, I'm drinking my own Kool-Aid here. I really do believe that this is going to be exactly what the doctor ordered in terms of leveraging this capability to drive those outcomes that I think businesses everywhere need. It's not like there's any industry that is immune from the impact of generative AI. It's not like there's any industry in any market that is not being forced to figure out ways to do more or better with less. And I think what's so unique about this capability is it accomplishes that, right? It used to be mutually exclusive: either spend more to get more or spend less and get less. Now, you can actually spend less and get more. Yeah. But you can't just plug in generative AI and wave your magic wand and stuff happens. You have to invest in that entire journey to identify what you have, to make that data actionable, to monetize it, to create access to it for your customers and team members alike, and then to deliver that better outcome. And I think we have a unique advantage here because of some of the investments we've made, both inorganically and organically over the last decade. Yeah. So that there's very few of us in the market today that can walk you soup to nuts across that journey, and, I believe that, you know, as we continue to put more firepower... So hiring up a significantly increased direct sales force, more channel support for sales activity, more consultative selling motion, and then focusing on which sectors we wanna be in, whether it's industry vertical domain expertise- Mm-hmm. ... or, geographically. I think we have an opportunity to capture meaningful share here, and at the same time, try and mitigate what has been a bit of an Achilles heel for us that I spoke to earlier, which is that- Yeah. ... client concentration risk. Yep, yep. I wanted to pivot just a little bit to WillowTree. You know, you've owned it now for close to a year, not quite, I guess, at this point. I'm just wondering how you think about it strategically. You know, when you, when you made the acquisition at the time relative to kinda where we sit here today, and what do you think has to happen along the journey to kinda get it back to where you would like it to be? Obviously, there's some external factors that have been at play. But as those start to abate, you know, maybe you can tease out what you think will be some of those positive outcomes around WillowTree. So perhaps not entirely surprising, I don't think the WillowTree acquisition was a bad acquisition, notwithstanding, the disappointing performance of the asset year to date. Oftentimes, it requires the rear view mirror and the opportunity to look back and say, "Yes or no, that was the right thing- Right. ... to have done," and not dissimilar from what we just discussed. You know, if you remember, when we bought LAI, you know, the first six months of our ownership, it was problematic, and that's an understatement. Like, revenue growth was down double digits, starting with a four in the first six months. And as we learned, unfortunately, in the last quarter of 2020, going into between when we signed and when we closed, they were already starting to deteriorate, and that wasn't shared with us until- Mm-hmm. ... we'd already written the check. A bit frustrating, as you might imagine, but buyer beware, it is what it is. From then till now, that business has, you know, doubled in size, both in revenue and profitability, in less than three years. Right. I think that's a pretty good performance, outcome. And as we just discussed, that is really the foundation for our AI capabilities. I think that will turn out to be one of the smartest, most strategic investments we've ever made in the history of TI. Not dissimilarly from that, I continue to believe that the WillowTree capability, then and now, is critically important. I mean, given the complexity, just how daunting knowing how to embrace generative AI offerings are, if you don't have a consultative selling motion- Mm-hmm. ... and, you know, the requisite reference ability and compelling examples of where you've successfully helped other large enterprise clients implement these tools, I think you ought not to be surprised if people say, "You know what? Appreciate, you know, your effort, but we're gonna go with Accenture, or we're gonna go with, you know, someone else- Yeah. ... who's a more well-recognized brand, that has the brand recognition and the, you know, vendor risk mitigation attributes, that folks are wanting to take." You know, you may remember during our roadshow at the IPO, and candidly, every quarter thereafter, up until we bought WillowTree, you know, as we talked about our design, build, deliver pyramid construct, where we were, I think, under-edged, where we simply didn't have the firepower. The bench strength was on that consultative advisory capability to, you know, start the journey for leveraging these exciting, compelling technologies, and that's what WillowTree was and still is. And whilst they, like we, as I just said, constrained in their ability to get on with it because of this uncertainty that continues to plague decision-making, when, 'cause I'm certain it's a when, not an if, things start to loosen up. I say that because, again, no matter what the industry is, someone in your industry is making progress against these ambitions, is finding a way to serve their clients using these tools more effectively. If you don't, your customers go somewhere else. It's as simple as that. All of us, forget we're in business, you're just a consumer now. You wanna procure a product or service from your favorite vendor supplier. You wanna do it when and how you want, whether it's voice or chat or email, any hour of the day or night. You wanna be able to buy it, you wanna be able to return it, you wanna be able to change it, you wanna learn about it, et cetera, virtually all the time. And it's not just Gen Zs or Millennials. Even we Baby Boomers are actually starting to get the hang of talking to a human being as a channel of last resort. It's an admission of failure- Totally. ... because historically it's been so soul-destroying- Yeah. ... to have to wait around for that. So we all wanna use these tools, these channels, in a way that automates things. It just makes... It's a better user experience. So any business, if they're not building an ecosystem that allows their customers to interact with them in that way, they're gonna get beat by their competitors. So I think we have exactly, as I've said many times now, what the doctor ordered, but we gotta be patient and continue to demonstrate our capabilities and our staying power. And when the tide turns, I think our Fuel iX offering will be flying off the shelves. Awesome. Yeah, I'm looking forward to that, too. I wanted to just talk a little bit about delivery cost and margins. You touched on the margins a bit, but it, it feels like the pressures on the delivery model a little bit are abating. Meaning, you know, like, early on, wage pressures, elevated attrition, like, you name it, there was a lot of things. I feel like that is less of the discussion today. So one, is that, is that accurate? And then two, can you also just frame again for us kind of where you stand in terms of that cost optimization for your own business? 'Cause it is, it is driving some meaningful margin benefits. You know, I guess it probably when you're talking about 10, 50, maybe even 100 basis points of differential, your commentary is probably accurate. But it has always, and I believe will always be a fight to the death. I mean, wage inflation is omnipresent. Yeah. But, do you have members of your team who join the business and a year later ask you for a reduction in pay? I mean, just think about it, right? Yeah. I mean, you come to work, you get good at it. Your expectation is, "I'm now more effective, more productive. I deserve to be paid more." Yeah. Like, there's not a single industry I know of that is immune to that dynamic. So wage inflation is everywhere all the time, and of course, there's some variability, in some markets, it goes faster, and, you know, whether it's surplus or shortage of particular skills. But the skills we want are the skills everybody wants now, right? It's software engineering, prompt engineering. They are always in demand, and there is no market I know of where all of a sudden they're just punching them out like cookies- Yeah. ... and you could pick them up for a discount. On top of that, as we just discussed, I don't have too many customers that show up, you know, and say, "You know what? Things have been so good for us, we'd like to give you a price increase." Yeah. They just don't do that. And even when we ask, even if we can show, you know, we're delivering this much more value, et cetera, et cetera, times are tough everywhere. So we have this constant price compression pressure and wage inflation pressure. And it's not just wages, although in the technology services industry, that's 60+% of our cost to serve, but it's utility costs, it's real estate costs. I mean, the list goes on and on and on. So it is a constant battle to find margin yield. So we have to continue to drink our own champagne, eat our own gourmet cooking- Yeah. ... and deploy these capabilities, tools, technologies, process excellence inside our business to mitigate that, so that we continue to derive that margin yield. And I think, excitingly, the very things we're purporting to sell to our customers that drive their costs down and improve their client experience is the very thing we can do for ourselves. And I continue to be quite excited about how our mix shift by moving away from more labor-intensive solutions, so that every dollar of revenue we derive doesn't require the same $0.60 of labor to fulfill, but it's $0.40 of labor, and- Right. ... the rest is either margin or technology. That's great. Well, we are out of time, as you can see, but thanks so much for the discussion. There's a lot of good things happening, and I'd like to say you called the bottom, but maybe not quite yet. We're getting close. Fingers crossed. Okay. Thanks for your time, again. See you again. See you again. Thank you so much. It's a pleasure.