Hey, good afternoon, everyone. Thank you for joining us at the 29th Annual Oppenheimer Technology Conference. Happy to have with us the team from Docebo. We've got Brandon Farber, CFO, and Mike McCarthy, VP of IR. Welcome, guys. Thank you for having us. Really appreciate it. For the audience, we're going to have a fireside chat here. We will take questions from the audience if you want to submit those into the conference portal. I will try to address those towards the end of the presentation. With that, Brandon, why don't we start? Great to have you. Look, I think most people by now know Docebo, but if you could provide just some quick background on the company, I think that'd be a great starter. Sure. Docebo is an LMS that over time has focused on what we call hybrid use cases. Most people, they interact with an LMS when they start their career. They think of an LMS as a simple onboarding compliance and talent development tool. Docebo from the ground up was always built for the purpose of not only training your internal organization but also your external training. What I mean by that is customers use Docebo to train their partners, train their customers, train their memberships and associations. Large sports organizations use Docebo to train their referees, their local coaches. Learning has turned into a bigger category than just internal training. Docebo was at the forefront because we had the vision of building a platform to handle multiple different use cases. When you think about the LMS category, it is quite fragmented, where there are a number of LMSs that were born that only do internal training. If you think of your HRISs, they are a simple bolt-on for your internal training. Then over time, there has been LMSs that all they do is external training. But when you get into the upper end of the market where companies want one tool for multiple different use cases, Docebo is best in breed and one of the few solutions on the market that can handle that use cases. In 2026, we expanded into the skills category, which we can talk about a little bit later. We are now on the journey of verticalizing. We have always been very horizontal in nature. Someone in sales would wake up in the morning and be on a call with a manufacturing company, government org, healthcare. We focused on the government verticalization about three years ago. That did require some certain product-specific compliance requirements and security requirements, and that led to FedRAMP certification in 2025. Now we are on the journey of verticalizing healthcare, which is a very niche industry that we serve. Those continue to add on for new growth levers for Docebo. Got it. That was very comprehensive. Really appreciate it, Brandon. Maybe taking a step back while also diving into some of those points you made, lots of strategic initiatives over the past couple of years after new management came on board, and we have got this upmarket enterprise push. You touched on public sector, getting into skills. Maybe just elaborate a little bit in terms of where we are in terms of progressing through some of those strategic initiatives and what was the thinking behind that. Yeah, the exciting part, if you are thinking about Docebo for the first time, is that a lot of these growth levers are new for us. We are at ground zero for almost everything we just discussed. From a government perspective, we started in the SLED space last year. This year is really our first opportunity in the fed space. We are entering into Q3 with the most pipe we have ever had. We continually exceed our pipe targets in government. It is a growth lever. It is a small percentage of our ARR base, but we are thinking that the growth rates in this vertical, our aspirations are much higher than what we put at Inspire, which was 10%-15% subscription CAGR. For government, you could think of multiples of that number. From a skills perspective, this was an acquisition that was dictated by us being in the enterprise space in 2025 and noticing that a lot of prospects were coming to Docebo and wanting an LMS with skills integrated and having one solution to handle both those different use cases. Truthfully, we were actually losing deals to competitors in 2025 in the enterprise space because we did not have a compelling story to tell from a skills. We went on the market, we got aggressive and said, well, first of all, we said, "Should we build this or should we buy this?" We saw that it is probably going to take three years to build, plus another year to go to market. When we think about a four-year run rate, we just could not wait that long. We went to buy a solution in January, and we are already seeing the fruits of the labor, where in Q2 we won two combined logos with learning and skills. Frankly, the really interesting part of that story is that both these deals we would have lost in 2025 because they had a heavy skills component. Not only did we win those deals, but we won at double the ACV than we used to because when you think about skills and learning, let us use some clean round numbers. If we used to sell LMS for CAD 100,000, we are also selling skills for CAD 100,000. Now that combined package is CAD 200,000. Very impressive. When talking about that skills piece, the 365Talents business, again, you guys just recently acquired it. Where are we in terms of integrating the product? Where are we in terms of getting that product across the sales teams for both organizations? I would just love an update on. Again, starting point seems like it is really strong, but obviously it could be much better once you get this thing fully baked into the core business. Yeah. From a product perspective, we have two phases of integration. Phase I is complete. Phase II will be done at the beginning of 2027, and phase II is when it is really a seamless experience. One login, two different modules. From a sales perspective, H1 was heavy on enablement, hiring some sales overlay, such as solutions consultants, generating pipe from our existing customer base. We are expecting more in Q4 of 2026, where we are going to see all the levers come together. What are those different levers? Number one, we are going to still sell skills on a standalone basis. If you have Oracle Learning, if you have Cornerstone, name your LMS, we are going to continue to sell skills. Number two, which we have already done very well, is when new prospects come in the door, we are going to sell them Docebo plus skills. And then number three is, and that's the one that is going to get better in Q4, is going back to our customer base. Got it. Then maybe getting more near-term focus. When we look at Docebo through the first half of the year, you guys have significantly outperformed our expectations. I'd imagine investor expectations as well. Perhaps give us an update on what you're seeing in the demand environment. Where are you seeing pockets of strength? Where do you still see potential room to improve as we get to the back half entering into 2027? Yeah. We came into 2026, and when we provided our annual guide, we went out with the following assumptions. We went out with the assumption that we've already saw strong mid-market in 2025. We extrapolated that into 2026 because we had good visibility in pipe gen and win rates. Government, we did assume solid performance in Q2 and Q3 in state and Fed. The one assumption that we were conserved on is we assumed essentially flat bookings growth in enterprise because we were tweaking the motion a lot, but we didn't have a lot of data points to say we've turned the corner. Come in Q1, we had a really strong Q1 in enterprise. We had a really strong Q2 performance in enterprise, and it's not only the performance, it's also the pipe gen, which is going to suggest that Q3 and Q4 is going to follow along the same trend lines. What's left to outperform on the Fed side, while we did embed in our guide a good Fed quarter, there is potential for it to be a great quarter. But the problem with Fed is that it's lumpy, right? You either win this million-dollar-plus ARR deal, or it gets pushed to Q3 of 2027. So we're in a little bit of a scenario where we know what we know, but we also don't know what we don't know because it's our first Q3 with our FedRAMP certification in rail pipe. Then on the enterprise space, while we did tweak up the assumptions, we do think there's still room for outperformance there. Got it. Understood. One of the questions I have been getting after your recent earnings last week is first half, obviously very strong. You guys are embedding in an ARR acceleration in the back half. I imagine a lot of what you just discussed is what is underpinning the confidence, but what degree of confidence do you have in that? Do you feel that that is sufficiently de-risked from a macro perspective? You mentioned some maybe deal slippage given government is highly unpredictable. I would just love your perspective on how to think about that ARR ramp. There are two ways to think about this. The first one I am going to go into is it is just math where we are comping easier quarters. If you think about Q3 of 2025, that quarter was the first quarter where we had real Dayforce wind down from our OEM relationship, which is going from 10% of our ARR to zero over a period of time. So in Q3 of 2025, they churned roughly 5 million CAD of ARR in that one quarter. In Q4 of 2025, we not only had Dayforce again, but we also had AWS. When I combine those two churns, that was close to 8 million CAD of ARR churn. While we are still going to experience Dayforce churn, we do not believe the magnitude is going to be at the same level as 2025. So we are going to be comping over easier quarters, and then when we combine that with what I mentioned is that we assumed enterprise flat bookings, and we are seeing bookings go up. When we combine the easier comps plus performance going up, it gives us the confidence that our top line ARR is going to start showing flexion. We reported a new metric starting in Q3 of 2025, where we report ARR excluding Dayforce wind down, FX, and acquired ARR. The reason for that was not to say that this is going to be a metric we are reporting for five years, and that number is always going to be above our top line. It was really to give visibility in the short term on what our underlying growth is, excluding the wind down. We do expect those numbers to converge sooner rather than later. If we are growing that number at 13.9% and top line is 9.5%, at some point, those numbers are going to have to converge, right? Got it. That is super helpful. I deliberately dragged it out before going into the topic [Non-English content] of the last few years. AI, obviously, that is top of mind for everybody. What are you seeing as far as customer interest in AI? What is Docebo doing as far as putting capabilities in front of customers? Yeah. Let me start off by our buyer persona. We serve an interesting buyer persona where our buyers are HR. We are not selling into developers or the IT. I would not say our buyer persona is at the forefront of AI, but they are very curious. Yep. What we are seeing is that a lot of our prospects and a lot of our customers, when we demo our AI functionality, they love it. They are like, "This is exactly the partner we want to work with. You are thinking of things we are not thinking of." At the same time, they are like, "I am not 100% sure how to use this just yet." We could get to that topic a little bit later, but we are seeing adoption increase. We are not in an industry where we are going to go from low adoption to very high adoption quickly. Some of the AI products that we already have is content creation, which has been in our product for close to a year now. We are seeing steady adoption, whether it is call your LLM-based content creation. Our virtual role play, which again, we disclose some usage metrics which are going up. That is your avatar-based role play scenario. Think about if you are a frontline worker at a QSR company, fresh out of university, and you want to get training on how to deal with angry customers. Sometimes instead of doing that in a real-life scenario, it is easier to do in a role-play scenario. We are starting to see in frontline workers a lot of interesting use cases for role play. Then coming in Q4 of this year, we are coming up with AgentHub and Knowledge Search. What does that mean? Knowledge Search is you are essentially going to be able to attach your LMS to multiple different sources, SharePoint, Jira, and Google Drive. Essentially, the logic is a lot of learning happens outside of the LMS right now. If I think about Docebo, we're still a small company, but a lot of my knowledge articles are on Google Shared Drive. So it's going to be really powerful for someone who's taken training in the LMS, but also needs to ask a question where the article may not live inside the LMS. That's coming out in Q4. Then AgentHub is the ability to create custom agents that will be specific to your workflows that will allow administrators to significantly cut down time on think of enrolling 10 different countries into 10 different courses. Manually it takes a couple hours, and hopefully in the future will take a couple seconds. Got it. Since we are investors in Wall Street, the natural follow-up question is always, well, you've got some of these capabilities that you're pushing out there. How should we think about the potential to monetize this? Are these just table stakes capabilities? Is it a new SKU? Obviously timeline wise, super early, so I don't imagine a lot of revenue flow through. But how should we broadly think about the topic of monetization? Majority will be through ad credits, so usage-based pricing, usage-based revenues. What we are seeing today and what we learned from virtual role play is that customers want initial allocation of credits to test it out. Yep. They want credits to do 50 role plays. From there, they're testing out the value because all this is brand new. We're in unknown territories. When a customer buys an LMS, they know exactly what they're getting. They've used an LMS before. The categories existed forever. These are features and functionalities that are new, they're exciting. We believe in the value, but the customers don't know the value yet. So we want to get this in the hands of the customers, give them free credits, and then start monetizing from there. So if you're asking me to start thinking on when do we see monetization pick up on credits a little bit more, we're probably thinking in back half of 2027. Understood. Makes a lot of sense. We haven't touched on profitability yet, but Docebo has made significant progress in terms of driving margin expansion. Where are we on that journey? Has the low-hanging fruit been plucked? Do you still see potential headroom there? Obviously, the natural follow-on to the conversation we just had is AI could be amazing. I'm sure you guys are seeing some great advances, but there's also a cost to it, both from an external and an internal perspective. How do you balance that? Is there still room for margin improvement with that as a potential overhang? Yeah. So we guided to about 20% EBITDA margins for 2026. Last year, we were at 18%. We have gone from zero to 20% over the past couple of years. Generally, if you look at consensus, it's about 22% for next year, and that's roughly aligned with what we're thinking of 2% EBITDA leverage per year until we get to our midterm EBITDA guidance that we put in our Investor Day deck of closer to 27%. Where we see that leverage coming from is, number one, G&A. So we're still at about, let's call it, 13% of revenues when we adjust out all the transaction costs. Our guidance is closer to 9%-11%. So let's use the midpoint of 10%, and we got 3% of the leverage from there. We're really focused on G&A perspective, keeping that flat from a dollar perspective. Of course, we have inflationary increases here and there, but really focused on getting leverage in G&A. R&D, you touched on it. It's probably an area where we're not going to get significant leverage. While we're seeing better productivity, coding faster, better than ever before, tokens are expensive, and the amount of stuff that we want to do continues to grow. If you think about it, this kind of goes against the argument of being able to vibe code things, is that the constraint for software companies has always been that we've never been able to hire as many developers as we want, and we've always had a three-year roadmap. We're still struggling to hire as many developers as we want. There's a lot to do. There's a lot of stuff we want to build. There's a lot of stuff our customer needs. We're going to continue to invest in R&D. Sales and marketing perspective, we're about, let's call it 33% today. That's another area where we think we'd get leverage out of. Our average sale price is going up year-over-year, which means quotas go up and our sales force becomes more productive. There's an embedded, more fixed structure within sales and marketing. You have your sales ops, you have sales enablement. Event spend is a little bit more fixed year-over-year. We do expect as revenues increase, we could gain leverage in sales and marketing, and certainly, AI does help in marketing and Revenue Operations in certain of those areas. That's really where we see the leverage coming from. Understood. Now, I wanted to circle back to another new initiative that you touched on earlier. You guys are verticalizing the business a little more, which is natural as you start to scale. You called out healthcare as your next vertical. I guess, how should we think about the investments needed to break into healthcare more meaningfully? Do you have some anchor points that could potentially bring you into those accounts to serve as references that hopefully speeds up the rollout versus maybe the cold start that you guys saw in public sector? Yeah, it's going to be a much faster return on investment in healthcare. We actually already have 5% of our ARR in healthcare. We've gained these customers not by doing anything special. It's more of inbound falling in our laps. What's also interesting is that we have a number of our top 20 customers that are healthcare customers. That just started our thinking of the opportunity is big. When we look at, we have the benefit of looking at all our data points by industry. We look at win rates, we look at demand, and we've known for years that our win rates are lower in healthcare because our product is not perfect for the healthcare segment. There are niche requirements from a compliance perspective. There's niche requirements from a workflow perspective that are only applicable to healthcare, and it only makes sense to build those if you're going to verticalize healthcare. For example, in pharma, if your platform is not GxP compliant, which means essentially every update you do could be verifiable in the future, you essentially can't sell it to pharma. Because healthcare is so niche, there's not a lot of players, the competition is lower, and it's a big space. There are a couple of learning companies that exist in the market. All they do is healthcare, and they're about the same size as Docebo. The opportunity is big. From an investment perspective, we've never been a company where we go crazy on investing before we merely have proof points. We're making an investment of, let's call it, 10 - 12 people throughout product, marketing, GTM. We're really going to go through that investment for the next 6- 12 months and see how it goes. We're going to increase and decrease investments as we get success or future data points. Got it. When I look at some of your KPIs, you guys have made a lot of progress on ARR per customer. How much of that is, like you guys have gravitated towards larger customers. How much of that is more of an ARPU dynamic where maybe you're attaching more to customers, you're obviously cross-selling better, you're upselling better. I would love to understand some of the motions that are driving that and whether or not there's further improvement there. It's a little bit of everything. Just to put some numbers into what you said, in 2020, 28% of our ARR represented customers who pay us CAD 100K or more. Today, that 28% is 56%. We've clearly moved upmarket. Today, if a prospect comes in and they have less than 300 customers or 300 employees, we essentially don't serve them. We essentially say, "Docebo's not the platform for you. We're not an SMB platform. Pass them on to a partner or let them find another LMS." We've really moved up to the enterprise market, and when you move up in the enterprise market, your platform becomes more complex. It's just natural. The bigger the company, the bigger the workflows, the more the customization. When your product gets more complex, it's harder for a company with 300 employees to use that LMS. They need a simple tool where they can onboard in a week and has two buttons to press. We moved upmarket, and another thing is that you're right, we've attached more products. We've innovated, but we've also acquired. I do think you should continue to see our ACV go up. When I talk about acquiring skills and doubling the ACV on the onset of a new customer, that alone is going to drive incremental ARR. Another thing is that, frankly, our ARR with Dayforce on our shared customers was lower. As that customer base churns out, we're seeing ACV naturally come up as well. Understood. On that customer tiering dynamic, we've gotten all the positive feedback on the enterprise market. I guess, how is the downmarket piece of your business performing? I guess it'd be more mid-market at this stage rather than the small and mid. Yeah, we still have about probably 20% of our ARR below CAD 50K, which we call more SMB. That's probably more of a, let's call it flat growth type of cohort for us. Lower gross retention, lower net retention. And why we've always emphasized enterprise and why we're always emphasizing as we move upmarket, our metrics will naturally improve is because we've always seen the net retention and gross retention significantly improves in that market. So we posted in our Investor Day a metric that when we look at those two different cohorts, there was about a 9 percentage points ARR difference on average over the past three years. So it's clearly, when we think about customer lifetime value in the enterprise, we're seeing that being significantly more. Got it. And at the start of the year, Docebo announced a 10% restructuring. I mean, it seems like it was a little different of a restructuring than maybe like what we're used to typically. Maybe just update the room in terms of what you guys were trying to achieve there. Where is some of that reallocation going? What are the potential benefits that you see once you guys complete that process? Yeah. So we announced that at the beginning of Q1, and it was completed within the quarter, so we're fully through that. So the history of Docebo is that the company was born out of Milan, and all our R&D was done out of Milan. So we had 100% of our developers, but we also had a large portion of our customer base in Europe. So that formula made sense. Over a period of time, Docebo really started to become successful in North America, and that revenue is now about 75% of our ARR. So we had a formula where 75% of our customers were in North America, some on the West Coast, 100% of our developers in Europe. And that formula, it just no longer worked, right? When you're serving a company, the West Coast, Silicon Valley, and something happens and they need access to a developer, well, in Italy, they're going to sleep at that time. So we really needed to change up the mixture and have a healthy portion of developers in Milan and Toronto. Milan will always be our center of excellence from an R&D perspective, but having a mixture of developers in Toronto and Milan for where we are today makes a lot of sense, and that was really the main purpose of the restructuring. Got it. Another initiative that you guys touched on, which I guess initially doesn't feel like it would be, I guess it'd be more dilutive to margins than additive to revenue, at least near term, is the idea of the forward deployed engineer. Would just love some thoughts on where you're putting those resources. How do you think that Docebo will benefit on the other side of it once you really get that program up and running? Yeah. As I mentioned, this is an area where we're putting investments in, but this is not an area where we're hiring 10 FTEs. It's more of closer in the lower single digits. The thesis is, I talked about it before, where our buyer persona is not as tech-savvy as some other buyer personas. So we truly believe building agents within an LLM significantly improves your experience if you're an administrator or power user. But if we unleash Docebo AgentHub without powering our users to create custom agents, I think we'd be a little bit foolish to really expect them to be able to code and build their own agents. Some of our customers, yes, they have access to maybe tens, hundreds, thousands of developers. But if you think about healthcare, who's strapped for budgets, I don't think they're going to have access to an internal IT developer. So the purpose of an FTE is to go in, talk to our top 20 customers, which a lot have already shown interest, and be on-site with them for two to four weeks, understand their pain points, understand where they spend the most time in the platform, and create custom agents that will allow them to significantly improve their experience. Now, last year, talked about on the conference call is that we will likely have agents that are applicable for a specific vertical. So if you're a QSR company versus healthcare, those agents that you need are likely going to be significantly different. So the reason why we want to have an FTE is, yes, they're going to go to our top 20 customers, but they're likely going to create agents that are very unique and something that we could apply to the remaining of our QSR customer base. Really, we're still thinking about how to properly monetize this, but if you think back to what I said on credits and usage-based, the thing that's guaranteed to get monetized is that once the agent is built, every time it's used, that will consume credits. Got it. It is one of those situations where down the line, there is sort of a circular benefit here. You are helping to grow your customer, and then those customers are somehow facilitating the innovation of your future product lines. Is this an area where you guys have permission from the customers to do this? Is there a possibility that they would frown upon you using some of these learnings to obviously facilitate success of some of their competition? Yeah, it is an interesting question, and we will probably have to learn. But if you think about SaaS, like today, there are customers who pay us for custom development or custom workflows. But because we are a SaaS platform, that feature is available to all 3,500 customers, even though that customer paid for it. That is just the concept of SaaS, is that once it is in your code base, it is available to all your customers. What will be unique about an agent being built with a customer is that every agent will be connected to data sources that are unique to their customer. For example, QSR number one, they will have Salesforce, ServiceNow. Let us think about another one. Zendesk. The agent may be doing something specific that requires to tap into those different data workflows. But we are not interested in what it is connected to. We are interested in the backbone of the actions it is taking. What we are thinking of really, the custom build that we are interested in is not necessarily the connectors, which is important for the customers, but really, what is that agent doing? Is it automating, looking through everyone's content to say, "You haven't updated your insider trading policy in the past five years." There are very different things and that is what we are interested in, is what is the agent actually taking, and that will be proprietary to our customers. Got it. One thing that we are commonly asked is, okay, what is the capital allocation strategy going forward? You guys made a big acquisition with 365Talents. You guys have done a couple of SIBs. Yeah. As we move forward, where is management's priorities in terms of use of cash? If you take a step back, we have about CAD 45 million of cash on the balance sheet. We have CAD 90 million of debt. So we're running about CAD 45 million of net debt with EBITDA projected to be CAD 55 million this year. We announced a CAD 70 million SIB, which expires near the end of August, and CAD 60 million of that will be funded through debt and CAD 10 million through cash. At the time when we announced our SIB, our stock was lower, and we priced it at about 10% premium, but all the premium is gone. A rational investor would probably think that SIB is not going to get filled, which leaves our current balance sheet as is. Where are our priorities right now? A lot of it depends on the variables. We're not a company that is very static in thinking. Our capital allocation changes on a daily basis. The number one thing is, where is our stock price? When we look at our stock price when we announce the SIB relative to the EBITDA we're generating, it's close or less than 10x EBITDA. So we're a believer in the company. We're seeing accelerating top-line revenue growth. We're seeing EBITDA margin expand, and we just believe that the valuation at the current market is undervaluing the story. If the stock price continues to be at those levels, we will have capital allocation dedicated towards buying back shares. From an M&A perspective, while we've done two M&As this year, Docebo is never going to be a company that is aggressive in M&A. It's just not in our DNA. We're going to take the next couple quarters, really focus on integration, focus on execution. Of course, opportunistic M&A you can never predict. If you were to ask me, are we going to buy a company in the next four quarters? Probability is low. That tilts capital allocation towards investing back in the business, which we're doing, and buying back shares. Got it. That is super helpful. With that, we are up on time. I'm fortunately at the end of my questions as well. Brandon, thank you so much for all your insights today. Mike, always good to see your friendly face on the other side as well. To the audience, thank you for joining. Thank you for participating. With that, have a good day. Thank you.