Awesome. Thanks. We're going to get rolling here, it looks like. So thanks, everyone, for joining us. My name is Scott Berg. I lead our enterprise software and SaaS research efforts here at Needham. Today, we have UiPath joining us today. We have the company's CFO, Ashim Gupta. Thank you so much for joining us today. I was just kind of thinking, I've done this with the company the last three years, but it's been a different person every year. Daniel, a couple of years ago, I had Rob last year, and now I have yourself. So we're running the gamut here, at least. But Ashim, why don't you start off giving an overview of UiPath for the one or two people here that might be not as familiar with it?
Yeah, so $1.6 billion, approximately, of top line. When you look at our company, we started out. Daniel started the company back in 2007. Really spent a lot of time just on the product. Daniel, the CEO and founder out of Romania, spent a lot of time just on the technology, so the core platform was built as an RPA platform. Incredibly strong, very contemporary, very open architecture, which is one of the reasons why we scaled so nicely, and then over the last five years, it's really expanded from an RPA platform to a full-fledged automation platform, and what that means is the full lifecycle of automation, from discovering automations and process intelligence to be able to automate simple tasks to the complex tasks of our customers. We have a global customer base of 10,800 customers in and around the majority of the Fortune 500 customers.
When you look at it, we've scaled very broadly as well. 50% of our revenue international, 50% U.S., across every single industry, and when you break it down, our core customers continue to expand very nicely. Customers between $100,000 and $1 million, they're expanding at 119%, and when you go across our base, the entire company, as well as our customer base, is energized by what the agentic wave can do. RPA is a great platform to build off of agents, and agents can take the workflows of our customers and really heighten the value and increase the surface area of what we provide for them, so $1.6 billion of top line, generating $325 million of cash this year, looking to accelerate that next year while stabilizing our growth rate as we reach a really nice scale.
So talking about your product is actually one of my favorite things lately right now. We'll get to that question next. But the first question is really around kind of management changes at the company. I mentioned last year I had Rob on stage. The year before that was Daniel. Daniel moved back into the CEO realm about nine months ago. As we think about the business over the last nine months, has anything changed from an operational perspective? My guess is there's been some differences. But has there been anything significant that you would call out?
Yeah. I think, look, Daniel's the founder and CEO of the company. He was the CEO for many, many years of the company, more than seven years with the company, and obviously for the full lifecycle, but seven years from the product lifecycle mix. You know what Rob did for us is Rob really brought a confidence in selling to the C- level. And that is something that really helped our team and gave them a great energy. What Daniel has brought back to the company, though, is really a balance between a large company and a small company agility. And so what that means for us is a couple of things. I would name three. The first is a maniacal focus on the day-to-day execution of how the company is running. If you build your own house, you're very careful about how people walk in your house.
And so Daniel really drives that level of execution. And so all the tactical things, from the details of how we compensate our employees to how are we engaging with our top customers, that is a rigor that we brought in. Second is we announced restructuring. I think sometimes when you build a larger corporation, you get away from your core focus, which is your customers and your technology. And you build a lot of middleware organizations that are in between there. One of the things that we've really done is lean that out. That has helped to re-accelerate our profitability and drive the cash generation. But it also helps growth. Because when you have less insulation from the customer, you're able to act faster and act more in their areas. And the third piece is around innovation.
When you look at agentic and you look at the pace at which agentic is moving, our company needs to move at that pace, and focusing on fast, focused decisions with deep technology, that's a third point of emphasis that Daniel has really brought back into the company.
OK. So we're going to focus a lot here on product. I'm sure some of the other go-to-market changes and whatnot. But the part that gets me fired up is around the AI agentic theme a little bit. My background's in internal IT management. Did that before I got my MBA. Been doing this almost 20 years now. So I've been in those product kind of environments. Enterprise businesses for the last 20 years have really gotten out of the business of developing applications. They brought in SaaS applications for a variety of reasons. We can debate if they overbought certain SaaS solutions. But companies have really gotten out of the business of trying to develop applications themselves. Fast forward to today, you have these AI models, these agentic platforms that are out there.
And I think the average investor, I tend to debate with, believes that companies are going to ditch these software applications to some extent and try to develop them internally. I still don't know why a company would want to hire an army of developers to do that. Because they're not product experts. Models change. And then you've got to blow it all up in six to 12 months. Your conference, though, helped solidify that from some of the customers that we spoke with and how you talked about your new agentic strategy. Tell us why these AI technologies aren't going to usurp what this RPA space is. And this is really about bringing kind of a full suite of automation together and pairing two complementary technologies.
Yeah. Let me give a background just on Enterprise Processes. And I came from a corporation, General Electric. And so I was the CFO of their division, as well as shared service organizations. When you look at a workflow, a workflow is developed in multiple steps. There are probabilistic steps, and there are deterministic steps. What I would consider a deterministic step is things that are very rules-based, that you can't get wrong. You don't want a 7% error rate. You can't take an impact of how an aircraft engine moves through customs. Or for one of the Fortune 50 health care customers that has started using agentic on our platform, they're not going to go through claims processing and put that in a probabilistic process. So at the core of the workflow are deterministic steps. Those are the areas where core automation works.
Now, in order to build around it, it's not building a model. It is building an entire ecosystem of how your user interfaces and codes those automations, how those automations are run, and most importantly, how they are governed, and so those core automations, none of our customers are saying that they're going to go and build, use a model that is built to replace any of those enterprise-grade processes that are deterministic, that are scaled, and that require the appropriate level of enterprise governance across it, so that's kind of step one. What agentic does for us is agentic allows the probabilistic areas that could not be met by the deterministic elements of our platform to be able to come now into an area where they can be automated, which unlocks incredible value, so let me give you an example.
How you process the hundreds of thousands or hundreds of millions of claims that are processing your system, entering data. You're not going to give that. You're not going to give agents or models access to those systems of records to do what they think is correct. But when you have to go and decide which of these records should I audit, that is a probabilistic step that you want to be able to understand. You want to be able to go and say, of these 50,000 areas, if I run it through a model, which has the greatest area of fraud? That step, you can't do it through a rule. You're limited if you do it simply through a rules-based engine.
So by being able to do that and interpret different data, whether that is written information through LLMs or being able to use other elements of AI, you've now expanded the scope of what's possible for automating that area. That isn't just in health care. You look at airlines. We had one of the largest airlines come and talk at our conference from Europe. They are super excited. And they've received incredible value from our RPA elements, our Document Understanding, and even our Test Automation platform. That was our core platform. What they're excited about agentic is how to transform a lot of the work that they're doing with customers that require not rules, but some level of agency to be able to understand how to interact and how to respond to customers. So those are the areas where we don't see it as cannibalization.
We see that as really additive. The last point I would make is when ChatGPT first came out, the number one question I got, which to Daniel was infuriating as an engineer, is, "Hey, why can't I just tell ChatGPT, write me UiPath's platform?" Two years later, you don't see that in the market, and the answer to that is then every piece of software, every Microsoft Excel, I could do the same action. So those things I think are more around fears than true capabilities. What customers are excited about is the openness of our architecture allows them to use our models, allows us to use any model in the marketplace and give them access to it, or for them to plug in their models into our platform to inform the workflows that they've already been automating.
To your point, I kind of touched on it earlier, that the changing models, I think, has been the biggest difference in this space over the last year, at least the questions I get. It seemed easy, quote unquote, to develop against, name your model, Claude, OpenAI, et cetera. But we're seeing more vendors like yourself in customer conversations say, hey, I want this orchestration and workflow that's here. But I want to be able to plug and play a model. Take it out today because today it's the right model. But six months from now, it's not. Plug in a different model. I just find something else works better there. Are you starting to see that from your customers want that flexibility, actually?
I think customers want both choice. And interestingly, they also want specifics. So I'll give you an example. I met the CIO for one of the largest banks in the country. One of the things they talked about, their security, they are not going after and certifying every single model or platform out there. They're going to limit the use of which models can be used. So now every bank may certify different models. So in order to scale, you need to be able to have openness. That's to be successful across an industry. But a lot of our customers for different use cases are using different models. For customer responses, they may decide to use a GPT-4. But they may decide to use LaMDA for something else, for handwriting or some type of specific document type that they're looking to process.
Our ability to sit on top of all the models, give access to that, that is a differentiator for us. Why, when you see our earnings transcript, if you hear the word Switzerland, why it's one element of it, we're not tied to any one technology or any one model.
A subtle part of your innovation with these Gen AI technologies actually came out a year ago with your Autopilot functionality. I thought it was almost ingenious because it could help with adoption of the platform. Because while I know the platform is developed for the end business user, I still think the average user needs a little bit of SQL scripting experience to really kind of understand it properly. Has that actually helped with the adoption of really kind of pushing the platform out to more employees?
Yes, very much so. I mean, in third quarter, we closed in second quarter, sorry, we closed one of our largest deals, which was agentic. But it wasn't with agentic in the way that our new products are being used. We actually used Autopilot as the front end for that. That is the example of blood plasma or blood plasma donations in which you're screening different donations. And millions would be rejected because they didn't have the ability to really interface or understand what the various donation parameters would be. That is an example where Autopilot, being a front end and also helping with the scripting on the back end, enabled that use case to come to life, not for a specific bucket to be scaled across the entire user community for that health care organization.
When you think of the agentic functionality in the platform, whether it's today or the next, I don't know, one to two years, how do you think about monetization? Is it going to be a separate SKU that you charge for? Is it really just kind of built into the base product, just trying to maybe help us understand how that may change the P&L?
Yeah. I think the first piece is we will monetize agentic. I think in terms of the method in which we're going to monetize, there are definitely going to be SKUs that are going to be launched. The question is, what is the right avenue for monetization? So we have over 1,000 customers sign up to our private preview. That is the largest number of customers that have signed up to any private preview in the history of our company. So the energy is there. And we're getting a lot of feedback from it. What we want to do in terms of our process is make sure we go through the customer feedback. We understand the pricing strategy. We introduce a set of SKUs, which we are going to be going into public preview with, and then deciding what's the right way to monetize.
Do you monetize the development of the agent? Do you monetize the consumption of the agent? How do you go through? How do you actually go and do that? One of the things is we want to hear the customer feedback before we solidify it. What we're conscious of, we've seen over the last year is a lot of promises of monetization that have not panned out, and a lot of that is because customer value or customer feedback comes back, and it's too hard to pivot, and you end up kind of just running into a wall. We really want to use the customer feedback to understand the best way to extract that value.
This question's come up a couple of times kind of historically is, how do you think about trying to report on your usage or metrics around customers using your agentic platform? You're clearly not there yet. It's too early. But do you think we can get to a period of time, I don't know, a year from now, two years from now, whatever it looks like, that you can actually kind of report around what the impact looks like? Because I think the narrative in the RPA space, as I mentioned, is these Gen AI technologies are going to usurp them, which I'm not a believer of. But do you think you can get to a point where you can actually have something reportable that really shows your impact in the business?
Yeah, 100% yes, and I think one of the things which is interesting that Daniel and the team are really driving is a focus on adoption internally. When you think about it, we have a multi-cloud strategy and a multi-deployment strategy. But more and more, I think the first step to be able to do that reporting is getting people to the cloud. The on-prem offerings that we've had, while very beneficial to our customers because it's allowed them to use and adopt and make us the incumbent while they're kind of working through their cloud roadmap, we want to get more and more customers to the cloud. You see that with some of the cloud metrics we showed at the beginning of the year in terms of just the growth rate of overall adoption of our cloud. Getting more people adopting to our cloud unlocks the telemetry.
With that telemetry, I think being able to report on that internally, report on that for our customers, and do the appropriate external reporting is within grasp.
So interesting. I'm going down a slightly different road here if you haven't noticed. But are you seeing customers actually make the migration to the cloud or have that conversation because they want to get to these technologies today?
Yes. Look, I think that different industries in different countries move at their own pace, so financial services, they move at a slower pace than manufacturing. Health care is a little bit more progressive than financial services but also moves at a slower pace than manufacturing, but we are a cloud-first company, and the majority of the ecosystem of software is predominantly cloud-first, so I think more and more emphasis on getting to the cloud and being able to unlock the latest features that we bring, that is a point of emphasis for our sales team and for our customers.
OK. Just moving to general go-to-market a little bit. I think if you look at the results, the company's results last three quarters, net new ARR has actually been down each of the last three quarters after having a little bit of a bounce back in the back half of last year. What's happening in the environment to kind of drive that right now? Is it macro? Is it execution? Is it something else? And what kind of what do you think changes that here going forward?
Yeah, look, at the beginning of this year, I think we talked very openly about the macroeconomic environment. We were owning that. But we also admitted the points on execution. I think when you're farther away from the customer, it puts you farther away from the core things that you need to do to drive your business. And what's ironic is it costs you more. People talk about investing for growth. I would say investing to be as close to your customer and surrounding your customer, that's more the right mindset than just spending money on go-to-market. So a lot of the execution areas that we feel has inhibited us over the last year, those execution areas we feel we've made a lot of progress on. We've reduced the silos within the organization, within some of the restructuring that we have done.
Second is we fixed some of the compensation plans that were honestly rolled out in a suboptimized way or something that created disruption. We've been able to quickly refocus that as well as stabilize those areas. So as we commented here in our last earnings, as we enter in next year, while first quarter, we'll see a little bit of the normal seasonality relative to fourth quarter, we really see the overall year as a year of stabilization of net new ARR as we fix a lot of the execution items that we've highlighted.
As you think about that growth algo, you guys have kind of talked about focusing more on your largest customers a little bit. Is that being able to drive more expansion within those customers or maybe just landing more of those large customers that can drive that expansion?
Great question. I think the first is quality over quantity. I think in the early stages, you can get carried away on the number of new logos you report in the quarter. But I always give this just as a metaphor. But you kind of don't want Joe's or Brenda's Pizza Shop. They're not going to scale automation or really drive agentic. So our focus has been on quality of new logos. And where we see that and where you see that in our metric is when you look at our customers greater than 100,000 or customers greater than a million, they're continuing to grow at a double-digit rate. While our overall customer base is standing relatively stable, the majority of the customers that we have shed, so to speak, is the lower-end customers, customers in the lower end of the segments that have a lower propensity to spend.
So our focus on large customers is both for better expansion and with a higher propensity to buy, which leads to better expansion over time.
Makes a lot of sense. Last question on the go-to-market side is the company announced a pretty tight partnership with SAP. I think it was your customer conference at Forward a year ago in 2023, kind of the unveiling of it at least. I guess how has that trended right now? And what are your expectations on that? I found it interesting given what they're going with their S/4HANA upgrade and how this can benefit, obviously, that migration for them. But should that, I guess, be a key partnership as we start thinking about calendar, I don't know, 2025 and 2026?
Yeah. So let's kind of talk about the partnership, and then I'll give a little bit about where we are in that cycle. It's a super excited part. We're super excited about the partnership. Two reasons. One is it gives us access to 40,000 customers within SAP, and the second is we essentially are inheriting their sales force to go to work for us. SAP will compensate them on the deals that they knock down, so that is just a great compensation framework that gives us overall go-to-market leverage over the mid- to long-term. What is the value proposition, so the value proposition is two-fold. One is when you move from S/4HANA over, you want to customize it less. That's a key goal for SAP. So to customize it less, they talk about it as a Clean Core.
To maintain a Clean Core, you can really do a lot of your customization through the automation that surrounds the ERP. Robots and agents are not as much worried about how the screen is configured or certain configurations. You can automate that process upon delivery of the ERP or as a part of that implementation. That value proposition, that is felt by our customers. We've already seen some traction and some wins for it. In terms of where the impact comes, we have to also accept that SAP is a massive organization. We're going to be in the enablement phase of that partnership today. They're going through their own restructuring, their own transformation. Daniel talks to Christian Klein quite a bit. We're tied into their executive leadership team.
But in our minds, kind of slow and steady to get the right motion on the ground and making sure that we are connected at the customer level, not at the management level alone.
My natural follow-up to that was going to be you just talked about getting closer to the customer. Partners don't always help you get close to the customer. But it sounds like you're trying to bridge that gap to be more of a unified or joint sales process with them.
Yeah. I think that's one of the areas that I've been focusing on in my expanded role and one of the areas I would say from an execution standpoint. Partners are absolutely critical for us to continue to drive scale. Our areas for partners is to be very clear on the segments and the areas that are best for their growth, the best for UiPath's growth, and the best for the customer's success, and I think meeting with those customers regularly, having joint business plans with those customers, those are just foundational items that we are putting in place, and we're super passionate about being closer to our customers and with that, being close to our partners because our customers are close to our partners.
We have about six minutes left, and then I'll open the Q&A to the audience. I guess as we start with financials, you also kind of talked about your expanded role as COO here. How do we think about the impact of those customers moving to your cloud model a little bit more? It's different than your typical perpetual to cloud migration. Your customers might not even see that much, and end users might not even see a difference because your perpetual is just your cloud product anyways, just behind the firewall. But from a financial impact perspective, will we see much difference in the next couple of years? Just didn't know if there's ebbs and flows to.
It depends on the customer. One, we sell something what we call Flex. So in some ways, we've sold a SKU that gives them the option of where they should deploy. What's exciting about moving them to the cloud is a lot of times the customers want what jump-starts their move to the cloud is the advanced features and functionalities they get just because we are a cloud-first technology company. So a lot of times, move to the cloud comes with certain upsells. They get access to our Document Understanding process. They get access to Test Automation. They get access to Communications Mining, which is really driving a good level of analysis of emails to understand where work is getting done and how to automate it. So that is one of the many benefits of cloud.
From an economic model, we still maintain our gross margins will be greater than 80%. We monitor, of course, 606 revenue recognition has different rules for cloud versus on-prem versus hybrid. So we continue to monitor that and be very transparent with the street and investors in terms of where we are in that progression and cycle.
You talked about the 606 aspects, which I still get. It's been almost four years since the IPO. I still get all those questions.
I still get all those questions.
I'm sure you do. I remember when you were talking to all the analysts pre-IPO because I think 606 was half of the conversation almost. But obviously, moving to the cloud should reduce the long-term kind of difference between ARR growth and revenue growth because they're more tightly tied together. Do you think in the interim, though, that could actually create more volatility around some of those contracts at all?
No, I don't think so. I think the metrics like, look, if you look at a quarter, you're going to see the impact between 606 in every one given quarter, et cetera. Last year, we had tremendous revenue growth, like north of 30%. ARR growth is stable. We've always maintained that ARR is the better metric to maintain the business. When you look at over the long term, like a trailing 12 month, trailing 24, you see those items converge, right? When you look at just today, just given some of the duration impacts of contracts that we talked about at the beginning of the year, we're seeing revenue trail underneath ARR. As we think about next year, we see that correlating closer as duration and just the macro environment is more stable, as well as our strategy. So for me, it depends on what time period you're looking at.
ARR is the metric that I think best represents the health of the business.
Sure. Never thought I'd get into this business to be a 606 expert. But the last few years, I have been. I actually started with Taleo back about probably eight years ago or 10 years ago, whatever it was. Anyways, the company's growth rate today is different than when you went public three years ago. It is for every company in my coverage universe for a variety of reasons. But how do you think about the balance between growth versus profitability today? And how do we set that model up maybe over the next couple of years?
Yeah, I think so to answer it in two ways. One is we still aspire as growth as our number one priority. It's still early. If you think about the number of automation things that need to be automated, just reflect on the firms that you've come from. How different do they look today and the bodies of work that are being done versus three years ago versus five years ago? I can tell you walking the halls of Procter & Gamble, there's tons of work that still requires to be automated. My friend is over there. And that's a customer that we have different connections with. So when you look at it, I think that growth is going to continue to be and innovation is going to continue to be a priority.
At the same time, at $1.6 billion of scale, gross margins greater than 80%, you can be profitable, and especially if you allocate capital correctly, and often what we found is putting the right constraints drives the right business discipline, drives the right level of focus, and any time we've been more focused, our execution on the top line has been better. Now, as you think about the model, we've talked about being greater than 20% operating margin and free cash flows. When you look at just what we've talked about for next year, you can see that we're there and more in terms of where we are, and you can see the stabilization of the growth rate that we've talked about being kind of in that low teens area.
So when you put that together, that's, in my mind, like just a formidable company at $1.6 billion and continuing to be able to generate cash and growing both on expansion as well as continuing to win new customers.
Last question for me, then happy to open it up, is we publish some checks. We check some companies all the time. But we publish some I don't know. I think it was a month or so ago. My takeaway from a specific customer we talked about is how our partner, excuse me, has talked about how the use cases are expanding out of the office of the CFO. I think the space has historically been tagged as great for the office of the CFO. But where do the expanded use cases come from? When you reflect today versus three years ago, my guess is you're going to say, yes, you're seeing much more usage out there. But how do these AI or agentic models really kind of help that expansion cadence in those other functional areas?
Yeah. It unlocks really incredible possibilities. I mean, if you look at supply chain, the amount of decisions that a human has to make that you wouldn't put just purely through a rules space, which supplier do I have the best chance of driving deflation? Pull those contracts. Those types of actions that are there, it just expands in a huge way what we can do for every function that is there. Traditionally, a lot of automation and a lot of software has been geared towards shared services, finance. What we're finding is even in the front-end operation, whether that's customer service transformation. LLMs give more intellectual capability to the software. And that in itself gives expanded use cases for any place that a human exists.
With that, happy to take questions from the audience if there are any, well, with that, we'll give everyone a few more minutes to fight the elevator.