Please welcome Investor Relations Vice President, Justin Furby.
Thank you. I love that video. In fact, I was actually thinking we could just end it there and go right to Q&A, if you guys are okay with that. Got a few laughs out of that. Welcome to Rising. Welcome to Analyst Day. Really thrilled to have you guys make the trip out. We've got a great group here in Las Vegas. A great group joining us on the webcast as well. So appreciate all of you. I know there's a lot going on in the market, and this week as well. So, really appreciate you making the trip out. There's a lot of energy in this room. We're going to bring it up, though, as we go through the course of the day. We're going to show a lot of product. We've come a long way just in the last year, so excited to show you.
If you took in the innovation keynote this morning, you saw some of that. We'll show you a little bit more for you all and give you a look into what we've been working on over the last, really last year, but in the last six months in particular, a lot of acceleration around innovation. This is what we're going to expect today. It's going to be about two hours or so. We're going to go across much of the leadership team, and talk a lot about innovation, and then how we're actually delivering it to customers, and what that adoption cycle looks like.
I think as we go through the course of the day, and we'll close out with Rob, we'll talk about, and this is a question we get a lot, what does that journey look like in practical terms when a customer goes down that path with us? So, this is my obligatory safe harbor statement. During today's presentation, we will make certain forward-looking statements. Actual results could differ materially. Please see our latest Form 10-Q filed with the SEC for a discussion of key risk factors. With that out of the way, it is my pleasure to invite on stage our Co-founder, CEO, and Chair, Aneel Bhusri.
I'm right here.
Oh, there you are. Where are you?
I was just watching to see how you were doing. Good afternoon. Thanks for being here. It's great to see so many familiar faces. Hope you all saw the keynote today and in the Q&A you can ask any questions about that that you'd like. What I'm especially excited to hear about, you're going to hear from all the product and technology leaders that are setting the pace for the new set of products. Zane's got a lot to cover as well. Talk about the market opportunity, the strategy we're executing against. Hopefully what you see is that there's a big bench of talent at Workday that's driving the company forward, and in many cases, new talent, because it's a new era, it's a new time, and we've got largely a pretty fresh team driving us into the future. I talked about the re-refounding moment in the keynote.
How many people were at the keynote? Everybody? Oh, some people weren't. God. Okay. What were you doing? No, I'm kidding. If it wasn't for AI, would I have come back to Workday? I don't actually know the answer to that question. I get asked that a lot. But once AI showed up, I felt like I needed to be back, and I had many discussions with my co-founder, Dave, about the future and what makes a great technology company. What lets a technology company pass the test of time as a company that people admire and that does good work for their customers? It's winning across multiple technology generations. We were a clear leader in cloud. I was part of what happened client server. But PeopleSoft didn't make it to cloud.
Workday made it to cloud, and now Workday needs to lead in this next generation of AI, and that's a large part of why I came back. Also came back, which is about driving innovation, but also came back to make sure we're refocused on our customers, refocused on our employees. I'm nine months in. So far so good. This is not just another product cycle. This is something like none of us have seen. Everybody has all these hyperboles around what AI means. But I've been around long enough to know I've never seen a technology like this before. Simply put, it's trying to replace the human brain. No technology in the past has done that. It's all been around increasing levels of automation. Do I know where this technology's going to go?
No, but I do want to have a say in where Workday heads with this technology as it continues to evolve on an almost a daily basis. Just to remind you, I'm not going to go into detail. You've been covering Workday for a long time. We've been at AI and machine learning since 2013. One of the reasons is in the founding principles of Workday is we thought data would be really valuable at some point down the road. We weren't sure how. It was before machine learning existed. But as soon as machine learning showed up in that 2013 time period, I was actually on the board of Cloudera, one of the pioneers of Hadoop, and I saw what was the start of something really powerful and magical. We just kept investing along that trajectory, and today, those bets along the way paid off.
We're still down to one data model. One data model means we can aggregate all the data across all of our customers and train the models in ways that nobody can touch. Not ADP, not SAP, not Oracle. They have their customers on multiple systems. On the HR side, no one has the scale of employees that we have at 80 million to begin with. One security model. You don't have to adjust security going from version to version, which means when you're using AI, it only has to traverse that one security model for whatever application it's using, one version of the system. Every customer on the same version, so when you come out with the new agent, you don't have to try to customize it between six different versions of legacy systems. One agent works on all of our customer systems.
That's where we are and lots of transactions, 20 years of trust, which is not something to underestimate, especially with so many new companies emerging. They don't have the data, they don't have the trust. They don't have the ability to actually drive the models the way we do. To get that trust with customers took a long time, but the way we translated is into this universal master service agreement that lets us use their data in an aggregate way, anonymous way, but that lets us train the models in ways that no one else can do. The other part of what we're doing is just understanding that AI is just very different. It's not about an end-to-end business process. It's about probabilities and reasoning, and I said this in the keynote this morning, you can't get a payroll 90% right.
You cannot get a close 90% right. They have to be 100% right 100% of the time. Reasoning can add an element that helps you move a process from step to step, but it cannot replace that process. When we can take what we know with our platform and tie it to AI work starts getting done on behalf of people and business, and it can rely on those guardrails. That really is our differentiation. There are a lot of spaces where AI can just run amok. Maybe it is in the creative space where you are just creating new content, you are creating new images. Not in the HR and finance world. These spaces are highly regulated. They have permissions. In some places, in the U.K., you track religion. In the U.S., it is illegal to track religion.
You cannot have agents that are just running across your data without any insight into the rules and regulations that run their business. When you get to the financial side, same thing. Those same security permissions, workflows, and biz rules, these are what customers rely on, and that is what makes for lawful use of AI. That lawful versus lawless distinction is really important to us. For those that were in the audience, you saw the ad that we did with Peyton Manning on the lawful versus lawless. In the spaces of HR and finance and ERP, that really, really matters.
There was that talk about SaaSpocalypse, which I never believed in, but now I think it is becoming clearer in our space that companies like ours, AI is an opportunity to leverage those business rules, leverage those guardrails, and build AI on top of it, which for better or worse means that SAP and Oracle are probably fine too. They have deeply embedded business processes and functions.
We will see if they can build the AI on top of it, but no one is going to rip out SAP manufacturing that I have ever talked to. No one has ripped it out for 30 years. Why would they do it now? I talked about the distinction about the deterministic rails. That payroll example is one I always come back to. What we need to do as a company, and part of our refounding, is this differentiation is take these deterministic rails.
Every business process has a start and an end. Some business processes have 20 steps along the way. A lot of that work between those steps are very mundane. They could be automated. There is a human involved. AI can actually move the process from step to step, and in some cases, just redefine that business process. That is where Workday comes in, to take the strength of those deterministic applications and now come up with the agents that can move work across these different business processes. The last part I talk about is operating with a startup mindset. In coming back, you heard this in my keynote, if you were there. I did not really have a guidepost of how do you do this when you have been gone for three years, a new technology emerges, and you come back.
I watched a lot of Steve Jobs videos, and I've said this many times as much as he was a great product leader, he was an equally great management thinker and leader. His thing when he came back to Apple was to think about Apple as a startup again, and that's what we've done. We've got the scale, we've got the security, we've got the platform, 11,500 customers, but we need to innovate like a startup. We've organized that way in how we're running the company. How are we operating differently? We're doubling down on what matters most, concentrating on AI, on our platforms. Our core apps are still important, but I'll just say, talking to customers, that 100th feature in HR or finance is not as important to them as these new capabilities in AI.
Having said that, we still have a lot of our development teams still working on those features because we do have a backlog of things we need to work through for our customers. It also means doing fewer things. When I came back, there were 50 agents. Now they're 20. 17 in production, three more to come down the road, focused on ones that really matter. That still sounds like a large number, but it's not when you look at the scale and breadth of our platform. Moving away from niche products, focusing on the ones that really matter. We are a leader in HR. We're a leader in finance. Let's work on agents that really leverage those capabilities and really help our customers at scale. We are moving much faster. Hopefully, you saw that today.
A lot of that work got done in the last nine months. I'm honestly blown away. I'm so proud of Gerrit and Joel and Yasmeen and Max and Gabe and Alet. What they've done in the last nine months, I've never seen before. Even in the early days of Workday, we didn't innovate that quickly. Part of it is just this AI technology allows you to do that. Part of it is the mindset of the team. We are working at speeds I've never seen before, and it's starting to pay dividends. You'll see that maybe not this year, or I don't know when you'll see it, but it's going to start seeping into our business as well, because these agents are truly very transformational for our customers, and they will generate a lot of revenues for us.
A couple examples. Workday Decision Intelligence built that organically, concept to GA in four months. Singapore Payroll, using AI, localized in just eight weeks. Before that, we thought it was a year project, and now it was done in eight weeks. 10 times faster than anything we've done before. The speed really does matter. I guess the other part, which I've mentioned a few times, is we're making the ownership much clearer. When you hear from our product team, Yasmeen owns Data Cloud, Alet owns Agent Factory, Joel is our Chief AI Officer, Gabe owns Platform, Max owns Applications. When I came back, it was very diffuse. AI was owned by all the application teams. APIs were owned by all the application teams. Frankly, we were moving too slowly. When something is being done by someone and it's 10% of their time, it doesn't get done.
When you're up against a startup, they're devoting 100% of the time to the activity. We had to do the same thing. Putting leaders close to the work, close to the customers, making them accountable, that's what we've set out to do. I'm going to quickly talk about the five ways to monetize Workday. I talked about the different opportunities, so I'm not going to spend a lot of time with this. If you were in the keynote, you saw the five use cases, so let's just turn it into monetization. Agents. Obviously, we're going to sell agents on a Flex Credits basis. APIs. If you don't want to use our agents on the front end or you want to use something like Cowork, by the way, Anthropic integrates into Sana. It doesn't even integrate directly to our base level APIs. It integrates through Sana.
Workday Data Cloud is a way if you want to do deep analytics off to the side and marry with a whole bunch of other data sets, we got the Workday Data Cloud. Workday Extend AI, if you want to build your own custom applications. Again, Flex Credits based. Then, of course, we still have our core applications that we monetize. So five different ways. It really does cover every use case. We're largely neutral on which ones you choose. Do we make more revenue on agents? Yes, but APIs are ones where we don't really have to have a sales and marketing cost, so that can be equally powerful. Workday Data Cloud, brand new opportunity that we really didn't have in our business plan. Same with Extend AI.
This model of monetization, still new to Workday, but it's coming together in a really nice way, and customers are embracing it. We still have work to do on the Flex Credits front to make sure that they understand how that works. In a few minutes, you're going to see them all in action. With Agent Harness, which we announced this morning, we get you that same trusted foundation to every agent. Gabe will share that in more detail. On APIs, direct way to access and monetize agents. Customers won't always start inside Workday. They happen in another third-party application. That's okay. These APIs are purpose-built for AI. I talked about the Data Cloud. Extend AI, that used to be Extend Professional.
As you know from our earnings reports, it's been one of the fastest-growing revenue streams we've had, and we're seeing customers build agents in a very short amount of time. They're incredibly powerful. We're not going to build all the agents that run at Workday. We don't need to. But Extend AI lets us participate in the building of those new agents. I do want to remind people that we're not stopping investing in the core applications. We are continuing to invest in those core applications. So you'll still see us adding new capabilities. There's new areas like financial consolidations that we probably need to do a better job of. So there are still business process applications, and anytime we build a new business process application, it creates data that feeds the model. So we're not done with that area, but we have such good products in that area.
Customers are more focused on the features around AI. There you have it, five ways to monetize Workday. Each one matters in its own, but together they reinforce a very open platform. Take what we built with Workday and use AI to make it better. With that, I would like Zane to join me on stage and we will get into all the financial questions you want to get into. Thank you.
Surprise you from behind.
There we go. There we go. Kind of what I did to
Go ahead. All right. Good afternoon, everybody. It is good to see so many familiar faces out there. Aneel, as CFO, I feel like I am obligated to tell all of you, as much as I know you like bigger numbers, bigger slides, that this backdrop was repurposed from a meeting we had yesterday. We are thrilled that you are here. This is not all for you. This was actually intended initially for our exec symposium, and this room was two or three times the size it is now, and we literally had hundreds and hundreds of CXOs listening to our story and talking about AI and looking to us to understand where we are with AI. It was an exciting time. It is an exciting time to be here at Rising. 17,000 attendees in person, 14,000 virtually, I believe. Just a great time to be here.
It is fun. It is fun to be back.
It is.
Yeah, it is just an exciting time to be back building product. I feel like a kid in a candy store right now.
Right. Right. For those of you who missed, I know a number of you got in this morning, obviously you will be able to have a chance to see the keynote, but we have got all the talent here to give you a private showing as well. With that, Aneel, you have been talking about market size and market growth, and for those of you who were here last year, you saw that we had a slide up, looked similar to this, except instead of $276 billion of TAM, it showed $188 billion.
So we are very proud of the growth here that you can see both HCM and Fins TAM has grown nicely, both through innovation as well as just the market growth. So that is approximately $176 billion as we calculate it today. The area we are most excited about is the $100 billion opportunity for us under AI and Platform.
For those of you who were here earlier, you saw the power of Sana, you saw the power of Workday Cloud, and our agent-ready tools that are really driving that $100 billion of opportunity, which is more than 2x what it was just last year. What this doesn't include are some of the interesting things that we're now getting into, notably ITSM and, of course, big announcement for those of you who missed it earlier today, with C1. I thought this is maybe a good opportunity, Aneel, to talk about the C1 opportunity and how you've been—
Sure.
—thinking about identity.
Yeah. I'd first start by saying, I've been back for nine months. I've talked to a number of customers. When you look at that AI and platform circle, they are looking to us first. I'm not one to exaggerate. I think most of you have known me for a long time. Our window is not forever, but we have a window 12 - 18 months where if we deliver our agents, we're going to be the default choice for our customers. They trust us. We already have their data. No startup has the data we have access to. They believe in the guardrails we have, and they want us to help them make their businesses better. So we have this window where I have yet to meet a customer who doesn't say, "We're looking to you for the agents.
Please deliver." So I just think that that $100 billion opportunity is there. It's ours for the taking. I'll just say like Data Cloud, I didn't really have that forecast into our business.
Right.
I think Data Cloud could be a huge new revenue accelerator. Getting to identity, I have been talking about identity for a long time. I was one of the founding investors in Okta. I love that space. I think there is a nuance that is lost on the market. HR systems are the source of identity. We have massive share in the HR space. Not saying I would do this, but if we turned off access to Okta to our HR system, they would be in a lot of trouble, which just tells you we are at the front end. We are upstream of all the identity systems. It is a very fragmented industry. HR is the source of identity. Now we are providing all the governance around it for not just humans, but for agents. We can work with other IDPs, whether it is Microsoft or Okta.
We now have that agentic layer that manages human and agent governance around all the things that are going to happen in the world of AI. It is incredibly powerful. It is going to take a while for the market to realize. It is going to take us a little bit of time to ramp up the sales. I can tell you, it is going to be very disruptive to the identity space, and we are now in the security space in a place that we have a right to win. I am very bullish, and you guys have all seen all the reports about agentic identity and security. We are now front and center in the middle of all of it. C1 is a very special company. You will learn more about that over time.
That is right. That is great. As we align with the growth and opportunity for AI, you can see the improvements that we have seen in this category on a year-over-year basis. We mentioned on our second quarter earnings call that we are approaching $600 million in annual recurring revenue with our AI agents and our AI portfolio, which is up significantly over three times what we shared with you just this time last year. While it is still early days, really tremendous growth in this area. It is not just the breadth of agents, and again, those of you who saw the keynote this morning, you can see we have a number of very exciting opportunities and agents that we are building. It is also the depth. If you look at areas like Sana, Extend AI, HiredScore, and Paradox, they are all growing nicely.
Three of them are over $100 million, and Sana is approaching that number very quickly. We are very excited by that opportunity and obviously the growth in our agent capabilities. As of the second quarter, we had 300 customers that had signed up for Flex Credits. As Gerrit will talk about, we have over 5,500 that are using a number of our organic agents already. It is really getting that flywheel started. Rob is also going to talk about the opportunity on a customer basis and how you see that flywheel build up with an average customer. We are excited by that opportunity in aggregate. Now, when we break that down and we look at our customer base, we have been focused on the dynamic with the customer base. Obviously this, as you all know, represents 60% of our trailing 12-month subscription revenue.
We see this as a tremendous opportunity to leverage AI. If you look at that opportunity, just over the last year, it was the second quarter where we saw AI become just over 50% of that mix between core and AI. 50% in the second quarter, we expect to end the year with more AI going into the customer base than just the basic core. As we look to next year, we estimate roughly 70% of that selling motion will be AI-driven and that revenue will be AI-driven. It is tremendous growth. Aneel, maybe just touch on some of the customers that you are talking to this week and how they are leaning in on the AI story.
Just last week I talked to Julie Sweet at Accenture. They are rolling out the Self-Service Agent to 400,000 employees. That application is based on Flex Credits. You can just think about that Self-Service Agent Joel talks about, as employee concierge. It is probably 30% of what it is going to be in a year, but it is already an incredible application, and I think it is the first killer application. I suspect many of our customers are going to deploy it. BMO is the other one, and they are well along the path and very happy. The customers are also very excited about the Developer Agent. They want to build their own agents. So are the service partners, by the way. I would not ignore them in terms of talking to the service partners.
The ability to build custom agents that they can resell to a whole bunch of their customers are really powerful. Then we have all the functional agents. Across the board, a huge amount of interest in what. Every conversation I have with a customer starts around AI and our agents.
Right. It is really that flywheel effect. If you look over the trailing 12 months, we had just under 900 customers that purchased at least one agent, and the uplift we got from that subset of customers was roughly 12%. You are going to hear from Rob where the idea is to get those agents out with those customers, and you will see how that can ramp significantly higher than the 12%. It is still early days. We are starting to see appreciable improvement and increases, and we are excited about what is still to come with the existing customer base. Which highlights the other part of the opportunity. 40% of our opportunity is still around net new and how we are growing, both internationally and how we are leaning in particular around the medium enterprise space.
Rob's going to talk a lot more about our initiatives there. You've heard of Workday GO, and really how we think about not only selling AI into those opportunities, but leveraging AI. You talked about around the world as far as looking at Singapore Payroll and all of the initiatives we have to localize our product and accelerate that roadmap. There's also the time to completion, the time to execute, and the time to get those products out there. It's an exciting time with medium enterprise. We've got a lot of initiatives, and we do believe that that's a key part of our growth with net new opportunities going forward. Again, a nice mix there.
Yeah. On the medium enterprise, a lot of the international market is medium enterprise. I happened to be in Europe two weeks ago. Met with both large and medium-sized companies. I was kind of surprised how far their thinking is on AI. I thought maybe they're going to be a little slower. Europe, in general, was slower on cloud. They're not slower on AI. They're embracing AI in the same way, and I think it's really exciting.
Yeah, and a great opportunity for us. We've actually, as you know, seen good success on our international markets over the last number of quarters. We expect that to continue. Then the focus ourselves, and Aneel talked about a startup mentality and culture. When we think about investing in growth, earlier this year, we talked about really understanding what that opportunity set is. Today's going to be more about the product set. We're eager to hand over to Gerrit and his team to talk about those products and get you excited about those products. It's everything from rethinking how we deploy and how we engineer those products to the go-to-market side.
Rob's going to talk about how we use AI and leveraging AI in our go-to-market capabilities, but also thinking about the whole adoption, whether it's forward deployed engineers and how we lean in with our customers. It's really around how they adopt, and that's what they want to know and want to learn from us.
When I think about what we are doing in AI, we have to be customer number zero. We have to be the best user of our technology, and we have IT, what we call a BT at Workday.
That is right.
A leader, Ronnie, who is totally embracing that. When I look going forward, there are a couple things that come to mind. I do not personally think about running the business for margin. I think about running it for growth. But if we run the business the right way, the margin should and will get better over time for multiple reasons. We get back to startup mode, we have to reduce layers. We just did a reduction. It was painful, but 2.5% of the company, and I promise you, it was not for cost cutting, it was purely the fact that in some places in the R&D organization, there were 10 layers between me and someone actually building code or writing code. That is not how you move quickly. Those individuals were complaining about how long it took for them to get anything approved.
We slashed a lot of those layers to get back to that startup mindset. Not about cost, but about speed, about ownership, about moving quickly. If we are a good user of our tools and other tools that are AI-based, naturally we are going to get benefit from margin. We are going to set pretty aggressive margin targets to drive us to use these AI tools and we have to eat our own dog food, or we should not be selling these products to customers. So far, the early results are very positive. But we are looking at the products across all of our software providers and how we can use them to use AI to drive margin. It may not be in terms of more reductions in force.
Maybe there are, but I think there's also the opportunity to just not hire in areas that you were hiring in the past and just leverage AI instead.
We've been doing that through the course of the year.
Yep.
With headcount, Aneel mentioned earlier in the year, our intention was to keep headcount relatively flat. Again, we want to be very thoughtful in where we invest and how we invest, as well as a company-wide thorough review on how we think about people, process, and systems all working together to just get better at what we do. A lot of opportunity there. Our last slide before we hand it over is, of course, around the framework. Much of this, we talked about it in the last earnings call, so it may be a little anticlimactic for some of you in the room. But we're obviously holding to our guide this year. We mentioned on the call that we expect to have 11% in subscription revenue growth into FY 2028. We still believe there's some upside there.
Aneel, if you want to touch on some of what you're seeing through the course of the week, but I think we've mentioned that a lot of this is about adoption and how customers are embracing the technology, and that's where we see, with the proliferation of agents that we're building, that there is some potential for upside into FY 2028 and beyond.
Well, I am hugely bullish and optimistic on what we are doing on the agentic front. Just before this meeting, I met with four different customers, and what they are frankly worried about is the token cost tied to those agents, which tells me they want to use the agents, and we have to get smart in how we work with them to make sure that the costs do not run away from them. That is a good problem to have, that is a better problem to have than they do not want or use our agents. They want our agents, and in some cases, they badly want our agents. So I am very optimistic on that. It is hard to figure out what that looks like in a year. I know it is going to be great. Where along the curve does it happen? I do not know.
I look at the new areas around identity, I look at the new areas around ITSM, and those are all, to me, upsides in the model. AI, I think we are in the process of catching a tiger by the tail. Hard to figure out when these agents, we have got 17 agents who are all young agents. When will they actually take off? When they take off, they are going to be huge.
Agreed. As we talked about through this conversation, and we mentioned already, we have obviously focused on non-GAAP operating margin and expect that to be greater than 33%. As important to us is SBC as a percent of revenue. You continue to see the decline. We expect next year to be roughly 15%, but it is something that we are focused on and expect that number to decline even beyond FY 2028.
So we will start talking more about non-GAAP operating margin, including SBC, and you can see nice growth here, and expecting over 17% next year with that important metric. We know it is one that you all follow closely as well. With that, we are also focused on how we deploy and how we think about capital allocation. It was just this time last year when we announced a $5 billion buyback. I was thrilled to see that happen sooner than expected.
Market conditions obviously helped in that regard. We also have $4 billion authorized for an open-ended buyback as well. So excited about how we are balancing that. The last line item here I know is very important to all of you, and that is free cash flow per share. We are focused on that as well, expecting $13 this year and $15 into FY 2028.
Yeah, I guess I'd say just on the agentic front, the agents, the APIs, Workday Data Cloud, this new consumption model is a beautiful model. It's a flywheel. Once it gets going, it really gets going. It's new to us. It's not new to us as executives or leaders. It's new to Workday, and it's new to you and how you look at Workday, but it's a far better model than a people-based, subscription-based model, which will continue for a long time. I'm super excited. I see the bills we have to pay to Google Cloud and AWS and Azure, and I think, well, I'd like to be in that business, and now we have an opportunity to be in that business.
We have to do it in the right way for our customers, but once you start using these agents, I've done it myself with the Sana self-service agent. You get hooked on it, and you've got to stop yourself because I realize I was probably burning through a lot of tokens. I do think it's something that people are going to underestimate how powerful this is going to be for Workday. I'm confident of that. I'm not one to get way ahead of myself, but I see it. I see it. When exactly it happens, I can't give you a date. I think it'll happen over time, and then it'll just explode as customers get real value out of these agents, and that's the key. They need to get real value out of these agents.
Great. With that, the most important time here is with Gerrit and his team because we want to have this be more you seeing the products in use. We would like to invite Gerrit up, and then Aneel and the team and I will be here later on for a Q&A. With that—
Thank you.
—invite Gerrit.
Thank you. We warmed up the crowd, huh? That's okay. Hello. Good to see everyone. Did any of you catch the innovation keynote earlier today? Some? Not a guest about, hey Mark, you know. I know, I saw you. Anyone else? Good. Okay, a few. Okay, I'm going to do just a very quick recap of some of that, for some who may have missed that. But we're going to keep it brief, because I know I stand between you and some pretty awesome product demos. So I want to have it going on a double for all of you. So AI is pretty incredible technology, right? And we all can agree to that quite easily.
The interesting thing is that on the enterprise side, you know, except for very specialized roles like coding, the breakthrough really hasn't happened yet, right? It really is mostly being applied to the old ways of working. You know, it's just making everything a little better. If you look at total company output, total productivity, the impact has not been as large as of today. This is actually where we think there is a massive opportunity for Workday if you understand, A, what drives that and why is that really?
What is preventing that from happening at scale? Understanding that when you actually change your perspective and look at it from a Workday lens, there is actually a lot of growth that Workday can drive from AI. It falls into two categories. What does every single company right now go through, no matter the industry, no matter the country. More or less, but in broad strokes. Everyone is working through the AI transformation. What are they all trying to figure out? How do they need to reconfigure their organizations to be AI-led?
How do their jobs change, their profiles change? What skills do they have? What skills do they need? What system sits in the middle of all of that? Workday on the HCM side. There is a very interesting narrative, and I think some of you are actually guilty of pushing that. That is that interesting thought that AI is just a squeeze, just an efficiency gain. Do more with less. That is all AI can do, right? You automate. You automate, and you automate, and you automate.
But it is actually a quite limiting perspective. I would say it is a real misunderstanding. Because the whole promise of AI is that actually you can drive greater productivity, address new market opportunities, serve your customers better. It is a story of growth. That is what every company is trying to figure out. Not every company is trying to be only more efficient.
How do their jobs change, their profiles change? What skills do they have? What skills do they need? What system sits in the middle of all of that? Workday on the HCM side. There is a very interesting narrative, and I think some of you are actually guilty of pushing that. That's that interesting thought that AI is just a squeeze, just an efficiency gain. Do more with less. That's all AI can do, right? You automate. You automate, and you automate, and you automate. But it is actually a quite limiting perspective. I would say it's a real misunderstanding. Because the whole promise of AI is that actually you can drive greater productivity, address new market opportunities, serve your customers better. It's a story of growth. That is what every company's trying to figure out. Not every company is trying to be only more efficient.
Everyone wants to grow. That is the big promise of AI. The opportunity that we have now at Workday is that on the one side, we are building that system of Work Intelligence that allows you to completely transform your workforce. Again, which every company has to do. Everyone tries to figure out, what do I need to hire? What skills do I need to build? Where do I employ my talent the best? The age of over-hiring is over. Companies do not hire anymore to grow. That was the old recipe. The new recipe is that you actually grow the company, and then you may hire. Maybe. Secondly, what is actually preventing all of those organizations from adopting AI at scale? No one has a recipe for how to actually do that, and it starts with very foundational primitives like security and identity and permissioning and guardrails.
I am sure you all read the headlines with the Hugging Face incident, of OpenAI model breaking out and hacking a website. That sounds really concerning in the enterprise context because there are so many things that can go wrong when you say enterprise. From the obvious things like paying people, to the really hard things, compliance when closing the books. This is what makes AI hard. What if you could solve that problem? What if you could have a platform which lets companies deploy AI at scale in a safe, reliable, and in a lawful way? Two massive opportunities for Workday. Just very quickly on the first one. Because there is, again, also this idea that it is only compression and compression and compression. Yes, if you get very specific in graduate hiring in computer science, yes, there is a compression.
But if you look in virtually any other function, the picture changes dramatically. Hospitality, healthcare, companies are looking for people to grow. If you look in functions like finance, what you all know, it is a very efficient function, but many CPAs actually are close to retirement. There is a huge talent shortage on that side of the world. Now we are looking at the new roles that AI is creating. Forward-deployed engineer. Every company right now in the tech space-ish is trying to build a forward-deployed engineer function. It has grown more than seven times over the last year in relevance. That job was practically irrelevant before. AI is in every job title right now, in all of the job postings, and most interestingly, outside of tech, marketing, sales, customer service. It is really changing the job profile.
The second thing which it allows companies to do is actually changing how they define how they work, what their workflows are. You saw today, and you are going to see later, Work Intelligence, which is really a strategic toolkit for companies to drive the transformation of work for themselves. I just want to quickly recap on the progress that Aneel spoke about. Because it is easy to forget how much we have accomplished and how transformative that change has been. Last year, we said we are going to completely reinvent the user experience of Workday. Today at Rising, Sana is the new Workday home experience. You saw it earlier in the quick screen video, a complete reimagination of how work gets done.
I give you one example because it shows so nicely what it means when you free people from the work about work stuff and give them time to focus on what is important. One example that you all may like, I like it, that most people like, is taking time off, PTO. Sana cuts the time down by more than 80%. If you look at our customers, that is equal to 1.6 million working hours saved every year. That is just one tiny example. Now, just think about if you redeploy that time to something which is productive versus managerial. It shows you how big that leverage function is. It is insane. We talked about the open platform strategy, and Aneel spoke about Workday Data Cloud and what it can drive. We have seen such incredible momentum with our AI developer tool chain.
We have trillions of rows of data in Workday Data Cloud already, growing 75% year-on-year. All of this is driven by customer demand, who wants to use that data for their own agentic transformation, for their own UI use cases, use it as context, bring it together with other data for analysis, all possible. Based on that idea of giving enterprises strong guardrails, we have had our open platform for AI in early adopter for four months. More than 6,000 applications developed, multiple hundreds of agents built with it. It really stems from that need that, yes, I want to have a frontier model, of course. Why wouldn't I? But I want to give it a safe context and guardrailing system so I can run it reliably and lawfully.
Last but not least, a year ago, we were just getting started with Workday Organic Agents. We had a couple, but I think it is fair to say not many of them have been in production yet. Today, we are at approximately 6,000 customers who are using Workday Organic UI in production. We just achieved a milestone of 1 million users for Sana's self-service agent.
That just shows you that incredible potential that we have in really reinventing the work across finance and HR. So bringing it back to the convictions that we have, right? How we think our portfolio strategy, why C1 makes sense, you may wonder, and what we are going to show you next. The key steppingstone to overcome is that AI needs to be trusted in an enterprise context. That is foundational. You would be surprised how hard that problem is to solve because, guess what?
Current guardrailing mechanisms, current way of thinking security, all of that is conceived from a human, right? From a human continuity or from something like a service account. No one has really thought about what it means in the age of AI, because an AI model can now act on my behalf and is doing something very different if it is acting on your behalf and doing something else, right? Yet I need to understand the security model per request at the right time in order to enforce it correctly. It is very important when we say that AI must be trusted, that we mean with that at all AI in an enterprise context needs to fulfill that guarantee of being reliable, auditable, and secure.
The second thing which I have just said, and that's driving so much of our momentum, this moment of redesigning work, really reconfiguring every enterprise in virtually all aspects, team, talent, organization, process, and giving them a system where you actually move from this age of copilots. That is really the limiting factor of standalone AI, right? It's that smart side panel sitting on the side, having no understanding of how to actually change the process itself because you just prompt it in the traditional way of doing work. Like I said, there are some domains like coding where this is just incredibly AI-friendly. But in 99.99%, that ain't the case. Companies work fundamentally different, and that redesign moment allows us to introduce these autopilots, and they are incredible game changers because you all think economically, right?
I would say roughly speaking, in economic terms, AI assistant equals to software spend. Autopilot equals to labor spend, right? Because it's actually doing something independently, behaving like someone performing that function. Last but not least, this idea of making AI a growth driver through your talent by actually putting the org and the workforce transformation in the center of that conversation. Like I said, every meeting I have had, be it with customers, prospects, partners, even before I joined Workday and worked for a very different company beforehand, every company was trying to map out the AI transformation, and every of these companies were starting with, "How do I actually reconfigure my organization and my talent around it?" That puts Workday in the center of all of that.
Work Intelligence, that analysis platform for recomposing work on a continuous basis, that's the brilliance of Work Intelligence and its counterpart, Decision Intelligence, that the enterprise is not sequential anymore, right? Something that maybe we have all liked, annual plan, annual budgeting, zero-based budget, all of that stuff, going to a continuous fluid model that you constantly redeploy and reassess. And your talent mix. That's Work Intelligence. I spoke about the autopilots, and recruiting is one, Sana Concierge is another, Sana Coach is yet another one. We are bringing all of them to market because this is where we are seeing the biggest unlock, and a Trusted Agent Platform underpinning them all. It's the platform that we are building on, and that's why we are so confident in giving the same platform to our customers.
Again, I want to address it because you may have that false contradiction that, should I use a frontier model or the Trusted Agent Platform? This is not how it works. This is like saying, "Should I choose a CPU or a cloud hyperscaler?" Right? You have to think of it as a value chain and a stack, right? We are using all of these models, and we are providing all of these models, including Jev from TypeSafe, recent innovation on the block, if you will. We are exposing them and giving them everything which they don't have, like the trusted context, enterprise guardrails, workflow awareness, domain knowledge in HR and finance. We are bringing this, which you saw in innovation keynote, into all canvases. Very important to us that Workday meets you where you are.
This is why we are introducing candidate search and application in ChatGPT. We announced today the partnership with Anthropic for Sana and Claude. That's how it all comes together in the AI ecosystem around us. I want to say just two quick things about C1. Because, it is not just about acquiring a company. For us, this has been, like Aneel has said, an incredibly strategic move for two reasons. One is, identity is fundamentally broken. Comically broken, I dare say. What is the source of all identity in any enterprise? Who knows who is getting hired, who is moving teams, and who is leaving the company? Any guesses? You're not a shy crowd, are you? You're usually having the sharp questions and the sharp minds. The HCM system, of course. That's the source of all of it.
All of the identity vendors are basically trying to organize themselves around it. But it has big limits. One big limit is there is a huge delay from hiring someone to provision access or from someone leaving and deprovisioning access. Big question in the compliance space. But there is even a bigger one when you think about AI. Who knows what the person actually got hired for? What's their job? What are they working on? Who are their teammates? If you know all of that, what can you actually automatically infer in access rights? And how intelligently can you automate them in a way no one else can? If you're a software developer, what you should have access to based on the true work graph of your environment. That is a huge potential.
Secondly, the reason why C1 is so special is because it thought about it not from the old ways of the human identity, but the task-based identity, permissioning work that's happening right now at a task basis. This is what makes it unique, and this is what makes it so incredibly valuable for AI because it solves that biggest problem, how do I do contextual permissioning for AI requests on a real-time basis? Again, something for the quantitative side. The estimation is that only, I think in two or three years, agents probably will outnumber humans 50 - 100 to one. That is a real growth opportunity because guess what? The physical world is limited. There are only so many workers in the world, and there is a limit to the growth, what it ever can be. But a digital universe, it's unlimited.
Again, when you think about your economic model, this is why this is so interesting, because you can have indefinite economic growth because you are not resource-constrained in the way you were in the hire-to-grow economy. C1 basically gives us the platform to be the security enforcement and identity layer across all of it. It also completes our world model of work. In the world of AI, there are just limited ways of creating real value. One is you can provide the very foundational base model. You can conceive it as evolution. That's model training. That's the best metaphor I have. It's like evolution. It creates something with a base intelligence. And then everything which you are building on top of it and around it is contextualizing it through tools that you give them, or you can call them harnesses, right?
Things you give it to act, data and context you give it to allow it, what is relevant, what needs to happen, what is the goal, and how to do it, and embed it in a real-time flow of work. Now with C1 and identity, we truly have that across finance, HR, and IT. We have an incredible TAM opportunity, and Zane and Aneel spoke about it. I am, as the product person, I have it easy, right? I am saying, as by the numbers, I am not a product visionary person. That allows me to say that I actually think our TAM opportunity is much, much bigger. Much bigger than this. Here is why. We have the opportunity to go into new areas that no company, not even Workday, could actually enter the way we can now.
Just think about employee service, all types of employee service delivery. Pre-AI, that was a question of expensive software and shared service centers. Sana Concierge and our self-service AI already deflects more than 25% of HR cases today and gives back our customers a 20%-ish gain in productivity. This is just the very beginning. With C1, you saw how we extend this into all of the IT requests. We are going to do this on all of the finance scenarios as well. This is going to be a hyper-converged AI service center that we are going to provide.
Secondly, we are going to really bridge the gap from work about work that get people actually frustrated and also gets in the way from doing value-adding work in their companies through autopilots, to a place where actually people can work all of their time and what it takes to drive their business forward, versus alignment, coordination, management alignment, reviews, et cetera. It is a massive simplification of work. If you heard yesterday, Yves Morieux saying, Yves Morieux said, basically, right now, companies are regaining control of what they were actually meant to do because all of that mundane work is being automated away. Talent acquisition is the best example, where you can actually see how that transforms in a complete agentified way. Last but not least, finance. Finance is an annoyingly efficient function. When you think about total spend of a company, finances is tiny.
This is not where the opportunity in finance is. How much money do companies lose through uncompliant spend? T&E is the biggest manageable expense line that exists. Huge problem. How much money is lost on incorrect payrolls? You would be surprised. The economic potential for us in finance is that it is not about making finance less, but actually having all of those financial controls being applied continuously and in real-time, and actually having that perfectly executing company at any second, not just at period close. I know I said a lot, and I am going to give it now over to the demo team, which going to show you examples of all of that. You got to remember, when you think about Workday wins in all ways.
Whether you are accessing our platform through Agent-Ready Tools, Claude or Copilot, it is our AI APIs on our Flex Credit consumption models. If you are building agents and you want to run them on safe enterprise guardrails, and you deploy them into Workday, you are sitting in our runtime, and again, Workday directly is able to partake in that value creation. Last but not least, if you are going for the full, and dare I say, the best experience for HR, finance, and IT, you are using Sana. You saw how transformative it is, and all of those three things come together in one agentic workplace. Again, that we partake in. This is really the key.
No matter what path a customer chooses, and maybe I am telling you a little secret here, there won't be a single clear-cut customer who is saying, "I am only doing one or two or three." Customers are doing all three simultaneously for various different use cases, and the greatness of our model is that we are part of all of them. I said I would keep it short. I hope I could share with you the most relevant pieces of Workday strategy. Now I would like to hand it over to Joel and Charlotte, who are going to show this to you in practice. Joel, Charlotte, over to you.
Oh, okay, like this? I love it. Thanks, Gerrit. It is fantastic to be here with you all. I joined Workday through Sana and now serve as the Chief AI Officer, and my thesis was really quite simple. The company that owns the workflows is going to be dominating the world of AI. Since we brought the sort of startup spirit and frontier of AI in Sana, with the enterprise readiness of Workday, and now rolling it out to thousands of customers around the world. Together, we are moving from AI that is simply responding to questions, Copilots, if you like, to agents that can safely do real work, autopilots. I will show you what that means for some of our customers and why I think this creates a significantly larger market for Workday as well. Workday's opportunity, of course, starts at the core, but it doesn't end there.
AI is helping us innovate faster across financials, payroll, recruiting, and planning. I think the larger opportunity is really transforming the work that is around all of these systems. That is really where we are focusing a greater share of our R&D. Most of enterprise AI systems today are sort of Copilots that sit on a side panel, answering some questions, maybe supporting you with individual tasks. We call that single-player AI. Our agents can coordinate and perform entire workflows. They anticipate what needs attention, they advise on the best path forward, and they can act within company policy and controls, can actually get work done. This expands our TAM truly beyond the software budgets into the labor and services spend attached to that very workflow. I think this slide does a good job at showing why this opportunity is so large.
Across each one of these categories, we already serve thousands of customers, more than 80 million users under contract. We are not entering these workflows from the outside. We already have the customer relationship. We already have all of the operating context, and every repeated workflow is an opportunity for our agents to do more work, and that creates measurable value for our customers. It also gives Workday multiple paths to expand within our installed base. I think this is really where Workday has a real advantage. We already understand these workflows. We already have the context across people, money, and processes to get work done here, and we operate the systems where all of the action needs to happen. This really enables us to have agents that do real work. They do not answer simple questions. They actually execute the entire workflows.
Customers get capacity to move faster, and Workday grows as more and more and more of these agents run through our platform. Let us just take HR and IT service delivery as an example. It is high volume work that crosses HR, IT, and multiple systems that our customers are spending tens of millions of dollars on outsourcing firms on. Today, this market still starts with a ticket. In May, we introduced the Self-Service Agent, as well as Sana for ITSM, and today we are bringing all of those capabilities together around our new offering, Sana Concierge.
This is a new way to deliver employee service across HR and IT, and I think it is really delightful to use. Employees can get help with onboarding, leave, payroll, access, IT support, all of that in a single place. As an employee, I just type what I want and Sana takes care of the rest.
Instead of these employee systems creating tickets, sending that to a person, Concierge is built to complete that request itself. It is the patterns that we have seen applied to customer service around the world, but now working for you internally. It understands the employee, it understands the company's policies, takes actions across Workday, but also all of the connected systems around it. When judgment is needed, it prepares a case for a specialist. With the history, supporting context, and the next steps already served there, a specialist can approve it. This is very, very difficult to replicate without Workday's context and processes. Customers get fewer tickets, faster resolution, and lower support costs. This is a radical change. We already heard 25% increase in resolution rates across Self-Service Agent.
We are only getting started there, and the tens of millions of dollars folks are spending on outsourcing spend here, we think we can make a significant cut, too. Also, employees become productive sooner. I will show you shortly what that looks like. For Workday, this also creates a much broader sort of premium offering for our customer base. It also creates recurring usage across the employee life cycle. I actually just saw a tweet before I came up here, and with someone about our new user experience and said, "Oh, damn. Will I start loving to use Workday now?" I think it is a pretty good proxy. People love coming back to our product experiences now. We are bringing a consumer-grade DNA to our product experiences, and customers cannot stop to use them. We have a single employee that now can instantly onboard to the company.
We have HR and IT service delivery that can run on autopilot. It's a really incredible experience, and I actually want to show you live what it all looks like. Let's start by an employee getting onboarded here. Historically, this would've been an employee jumping across dozens of systems, creating a bunch of tickets, escalating that to humans, a human resolving the ticket, and a lot of back and forth. Here you can see the onboarding experience with Sana. It's all set up from day one. Sana looks at all of your systems, what you would be working on, and so on, and sets up your plan. Here, for example, you can see the Sana Learn experience has created all of the courses and the training sessions, already scheduled them for you and what you need to get done.
Sana's already gone ahead and scheduled the meetings with all of the folks that you need to meet your first week, your manager, your direct reports, and some colleagues that you will be collaborating with. Also, through Sana for ITSM, as you can see here in Concierge, it's automated all of the setup of all of the tools that you will need access to. On day one, this is all set up for me and creates a very delightful onboarding experience. But I can also just go straight to the Concierge experience, and this is literally a concierge. I can type in whatever I want access to.
For example, if I need access to Claude, Sana will route that, look at the approvals required for this, whether it's safe to provide me access, and it looks like this is all good, so I can just hit Submit, and Sana will give me access to this system instantly. That, which would've historically taken a lot of back and forth, created tickets, being escalated to a human, was now run by autopilots. That's what we mean by autopilots. Autopilots can actually get the work done for you. But we've also been really pushing on the user experience. What we've been working on here together with the team at TypeSafe is Dynamic UI. This is instant UI. This is UI that's literally generated in hundreds of milliseconds, and Workday is the first company on the planet to bring this to the market.
Here it means I can write something like, "What's my first paycheck?" Sana will reveal that for me instantly. I can type a query like, "Who is Victor Miller?" It's so fast, and it feels so obvious that you barely notice when it's being generated. Here I type something like, "Who's Victor Miller?" It pulls all of the details for me instantly. But it can also create these richer components, like a calendar view. If I jump in here and I want to schedule a meeting with Victor, it sends that out. For frequent Workday actions such as filing a PTO or a similar task, Sana generates the calendar schedule, pre-populates all of these different fields so it all looks good, and then I can send out that request.
So as you can imagine, a lot of companies are building conversational interfaces to solve this task. For very few tasks, that's actually the most efficient UI. UI-augmented AI is really powerful, and we're really excited about the product experience that we've created here. Since joining, Sana has entirely reimagined the user experience. We brought a lot of new agents to market, and I couldn't be more excited about this progress. To share a little bit more about what we've been working on, I want to hand it over to Charlotte. I'll see you soon.
Thanks, Joel. That was what our new experience looks like for our employees. What we need to also focus on is how we're looking at new decisions about the workforce itself. Right now, I'm going to play the role of a CHRO, and I want to review our hiring plan for the next quarter. So I simply ask our Decision Intelligence, once I do a quick refresh. Sorry, everybody. Live demos. So I want to review my hiring plan for the next quarter, so I'll ask Decision Intelligence to help me understand what's happening. The current plan includes 50 new hires, but I need to know if that's what's right for the business. Do we still need those 50 new hires? I can ask Decision Intelligence to understand based on the work we know about the team.
Now it uses our Work Intelligence capabilities, understanding the tasks and the skills that we have within our workforce, and identifies the work that can either be automated or augmented by AI. From here, it's going to find out any duplicated work, underused capacity. Skills Cloud will show the capabilities of the skills that we already have, and with external connectors, we can bring in third-party benchmarks as another source of context. The answer is we don't need 50. We actually need 42. That will reduce hiring costs without relying on a simple headcount cut. So I ask Decision Intelligence to update the plan. But before acting, I want to understand the broader impact. Decision Intelligence actually recommends that we even fill nine of these roles internally.
It searched my entire global workforce for suitable employees, the employees that have the skills we need, the experiences, and have even indicated those career interests. Now, this reduces the need for external hiring. It creates meaningful growth opportunities for employees, and my plan now actually reflects the skills, capacity, and the actual work. This is far more precise than a traditional headcount plan. But every workforce decision has a financial impact. So as a CHRO, I decide to share this plan with my partner over in the CFO world. We'll now move into the FP&A experience once this is shared out. So I share this with my financial advisor. My financial advisor? I share this with my CFO, and I'm going to be ready to review this. However, from a CFO perspective, this is the Decision Intelligence is bringing the same data.
There is no copy and no sync between HR. HR does believe that we can fill these roles internally. But from a financial perspective, I want to actually understand the impact that this will bring. From there, I notice that we see that we have faster internal hiring recommended, and this could mean faster revenue growth. So I ask that same Decision Intelligence about, if we fill these roles internally, what is our revenue potential? It is going to model that outcome. I see start rates, ramp dates, and Finance is now using a model that is no longer just cost data. This information includes people, skills, it has got the ramp times for the different roles, and it is even bringing in pipeline data and demand data from Salesforce. Decision Intelligence analyzes all of those inputs.
It shows that we can actually support a higher revenue target, and this target is based on roles that we can actually fill. It is not just based on hoping that we hire faster. Now I want to look at the information behind this. HCM provides the skills and the interests, but I can see where the hand-offs start to happen. I see Peakon is providing the sentiment insights and confidential comments, allowing me to better understand my employees. Financials has brought in the actual costs. Salesforce provided that pipeline and the demand data, and normally this type of amalgamation would take weeks and several spreadsheets. Now I have one, with the sources actively identified and clearly traced. I now have a plan that my business can execute. Decision Intelligence does not write back anything on its own. It waits for me to review and approve the changes.
Once confirmed in the working plan, the approval train will always be visible. All of Workday security and approvals remain intact. So what did we just do? We analyzed the data, we made a decision, and we recorded it. That is the value of owning the full decision loop. But as AI helps company move faster, financial controls must keep up, and traditional controls often test a sample after the fact, and by then, the damage may already be done by an issue that was caused. Financial test suite continuously tests transactions as they happen.
Here I see it is examining the work completed by people and agents, and in this example, it reviewed 90,000 transactions. But it has identified a small number of risks that do require investigation. Let us start with procure to pay. Here I see various tests covering several payment risks. These tests include duplicate invoices and cutoff issues.
We can also now evaluate actions taken by the agents themselves. In this case, this test reviewed more than 400 actions by an AP agent. However, these same controls applied to people now apply to agents. But I want to really investigate this payment change to bank account details after they were changed. This is where I noticed that there were some suppliers that changed their bank details shortly before payment. The agent identified the risk, placed the payments on hold, and now I can either remove the hold, cancel the payment, or investigate further. To do this, I can see the transaction lineage. This shows who approved and who initiated each step. From there, I can quickly see what changed, and I can also confirm whether it was within our policies.
Based on this, I can decide I might want some more coverage for these high-value financial transactions. I can go to my test library, where I am able to see tests that come from both Workday and our third-party partners like KPMG. KPMG here is providing a three-way match exception detection, allowing me to do that as an AP cutoff. However, based on what I just saw, I really want to focus on building out a test for a segregation of duties. Now I see a test is provided by Workday that will check whether or not one person can change bank details and pay. Definitely something we shouldn't be allowing. That is a combination of permissions that introduces significant risk. I ask the financial test suite to build out this and configure this new test.
Before publishing, I will be able to review the methodology that was used in order to build out this test, and I can even see a preview of the people and transactions that were flagged as a result of this test. Now, I notice that there are some responses, but the agent also anticipates and advises some changes we can make to this financial test that will improve the outcome. I accept the recommendation and rerun it and see the new results. Once I am satisfied with this test, I publish it, available to the financial test suite. Already within this, we see the results, and it immediately begins running across all of our transactions. This makes continuous compliance possible before an external audit. Now let's look at a very different process. Talent acquisition. This makes Workday a major gateway between employers and talent.
However, as recruiters, it still requires significant administrative effort. This includes screening, sourcing, scheduling, and all of the related follow-up. The talent acquisition agent changes that. Now recruiters can focus on the people and the business of judgment. Here in our talent acquisition agent, we see a new role that was created that didn't previously exist, such as our AI safety engineer. Normally, this information would wait until I logged in. Like most recruiters, I am managing a lot of different things. The talent acquisition agent doesn't wait for me. It continuously runs and scans every open requisition. Now I see that it's flagged a potential hiring risk. The AI safety engineer doesn't have enough candidates in the pipeline to reach a potential target hire date. The agent diagnosed the problem before I even looked at the requisition.
It's found out some previous applicants and members of our talent communities that I can reach out to. These could be silver medalists or the second-place runner-up in a previous requisition. Or we can even utilize external partners, such as Findem and hireEZ, as a sourcing opportunity. Now the agent has drafted a potential reach out to 15 candidates. At this point, I can review and use my judgment to approve it. This is hours of sourcing saved and now becomes more focused on a review. After I approve that outreach, I see that applications have started to arrive. Very exciting. It helps me identify the strongest candidates faster. I will be able to see and understand why a candidate is a great fit. This is bringing together our HiredScore, Paradox, and recruiting capabilities all in a unified experience.
But importantly, I also see the gaps and why they might need more attention. However, I think this candidate looks pretty good, and so the recruiter is still going to be responsible for the decision. I move this individual to an interview, but the agent has made this decision easier and better informed, and now the agent starts the interview process. It checks availability from the candidate and the hiring team. It sends the invitations and handles any of the rescheduling. That removes a large amount of administrative work. So during the interview, it means we can focus on the most important thing, the human connection that still matters the most. So we start the interview, and when it's time for the interview, the agent is there. Doing the interview is the actual fun part. Writing the feedback, that may be not as exciting.
The agent steps in with pre-populated responses based on transcripts and notes. I am not starting from scratch. This gives recruiters and hiring team the time they need back in their day, so they can focus on their teams and the interviews and removing that administrative burden. This interview then becomes another signal about the candidate. So from here, I am able to see and understand more about the agent's signals and what they are using for candidate rediscovery. It might decide to rediscover this candidate for another role in the future, should we not move forward with them. The organization also remains in control of all of this. In the control plane, this defines what the agent can and cannot do. It also defines what must wait for a person. This means the agent moves quickly, but within the boundaries that are set.
Now, we have transformed the recruiter experience, but we have also made applying easier for the candidates themselves. People are now discovering jobs in entirely new ways. The application experience should meet them where they are. Today, I am going to use ChatGPT to look for a new role. Little awkward with my entire leadership team currently watching me to apply for a new role, but we will do it anyway. The candidate profile now gives me one identity across Workday-hosted applications. I no longer need to create another account for every single employer. I verify my identity through a quick text exchange. Sana then surfaces roles based on my skills and who I am as a person, and the experience that I might want to consider. From there, I can see how well my skills fit these particular roles, but also where I may have gaps.
At this point, I decide whether or not I want to apply. I answer any of the inline questionnaires and the application experiences right in the flow of that work. And then from there, I am able to directly submit my application. There is no longer a long form of fields and candidates and passwords to remember. This ultimately can help strengthen candidate pipelines by reducing candidate drop-offs. So you have seen a lot, but I would like to pass it over to Joel to close out our little section.
Thanks, Charlotte. Wasn't that really good? I mean, you just saw how we're completely transforming the work across all of these Workday processes. As we do this, we drive tremendous efficiencies for our customers. I also think these are many illustrations of how this is opening up entirely new sources of TAM for Workday. In recruiting, for example, as you just saw, we are completely changing the economics of recruiting. Today, we have over 400 million applications on the Workday platform just last year. This creates an opportunity for both Workday and our customers, where our customers can continue owning the relationships and all of the hiring decisions, but the agent takes on more and more and more of this work. The more work that our agents take on, the more our opportunity grows.
It's no longer based on the number of users on the platform. This opens up a labor pool of more than $100 billion just on the recruiting space alone. These examples represent a much broader roadmap for us. Our agents can anticipate, advise, and act across many workflows, and over time, they take on more and more and more of this work. Of course, it starts with these individual tasks, but then it expands into the continuous coordination across complete workflows. Where we're focusing Workday is where we have that differentiated context, where work happens repeatedly at scale, and where that customer outcome is measurable. Workday gains new products and more platform consumption as a function of this, but we also get deeper engagement inside of our core applications. Today, we showed you some of these examples of where Workday can transform workflows directly.
I'm a very simple man, and my thesis here really is we own the workflows, we build agents that can deliver on these workflows, and we expand the TAM to take on labor as well. It's an incredibly exciting time to be building AI at Workday. Our customers also need an open platform that preserves this governance. I want to hand over to my good friend and our CTO, Gabe, to show you what this looks like across the full ecosystem. Over to you, Gabe.
Hey, everyone. It's good to be with you today, and I'm excited to dive into some of the coolest tech that is really transformative. It's such a special time to be in the industry right now. Gerrit hit a point that I want to pick up, which is enterprise AI only works if it's lawful, not lawless, and that means running inside of enterprise rails versus around them. What happens when something like that could go wrong? Why don't you take a sneak peek at some of the creative?
[Presentation]
Funny, right? But AI that's almost right is wrong. It's true, right? You can't get payroll right 99% of the time. It's got to be right 100% of the time. This is true in a lot of different scenarios in the space of HR and finance, and that's why Workday's role in this transformation is so critical. It's also why Workday is evolving from a system of record that you log into at your desktop towards an open platform that you can build on and that you can easily extend. We call it Workday's Trusted Agent Platform, and it's differentiated on three core points that drive our strategy and drive our platform business. The first is that we're open by design. This is a key to our monetization engine. I think Aneel and Gerrit both mentioned this, and we'll talk more about it.
The second is it's backed by builders. This is important because everyone today is a builder. Natural language has democratized the art of engineering these systems, right? We want to create that product-led ecosystem in flywheel. The third, and this is really critical, and this is something only Workday can do, is we are lawful by default, and this is our unmatched competitive moat, and we're going to cover each of these. Let's start with open by design. If you remember nothing else today, I think this is the one to remember. The exact path Gerrit mentioned, this is the one that lets customers connect their own agents straight into Workday. Again, Gerrit talked about this, but it's really important. A year ago, we stood up at Rising, and we promised an open platform. Today, our Trusted Agent Platform makes that real.
What does that mean? Headless access to HR and finance data. Every capability required to do that is now live in GA. The cool thing to me is we're customer zero for this stuff. If you take a look at the demos that were in the innovation key and all the stuff that Joel just shared, all of that is powered by all this headless technology, right? We're building on top of the same stuff. But it's also a major monetization shift for us. This is where we can create a brand-new product-led flywheel that's going to drive platform consumption and really extend the growth of the business.
Look, we don't have a lot of clarity on what this looks like yet because it's still early days, but we know there's tons of demand out there for folks building, whether it's personal agents or different ways of driving the headless business. It's really a big demand. To walk you through how this works and some of the challenges, frankly, that we see with this, we're going to walk you through a company that we've set up here, Acme Industries. I want to be very clear, this is modeled after the very real Fortune 50 customers that I see on this headless journey. Now, meet Maya. She's an Engineering Director for HRIT at Acme Industries. She's going to do a very common task, right? Building an internal mobility agent. Very common thing that we see from our customers.
She could do it the way that we see a lot of customers doing it. Own stack, headless. You're going to need a few things, a way to pull her open job recs, a way to move people between different roles. All of this, critically, and this is the part I want you guys to pay attention to, is it's got to be scoped to what Maya can see and what Maya can do. There's a couple ways to go about this, and before I show you the right way, I actually think it's more instructive to show you the wrong way first, because this is where I think the market has it wrong. I know certainly that some of the customers I talk to have it wrong on this.
It's wrong in the sense that people think it's safe to build agents on custom MCP-engineered solutions or data that's extracted out of Workday and put into some external data lake. I want to show you guys, this is not safe. This is actually causing a lot of problems. It is lawless, as we say. So let's show you headless the lawless way. We'll move over to the demo here. What we're going to start with is here, we're obviously in Claude, so we're going to think through this internal mobility agent. But first, I want to show you this is the pattern. Acme Workday MCP. This is not Workday. This is built by Acme. By the way, this is a true story, okay? I have lots of customers who are going about it this way.
Think of it as like they're creatively taking a way around our guardrails and our path here. Well, a lot of them don't mean to do that, but they are doing it. If we take a look at what's inside of this MCP server, we got some read-only tools, we got some write tools, should be enough to execute what we want to do. So let's just try out some of the gestures required to build this agent. I'll paste in, "Share me my open job reqs." Or sorry, "Show me my open job reqs." Now, the word my is doing some work here, right? My open job reqs refers to Maya, okay? Now, you're going to see we get 30 open job reqs back. Okay, so that's first part. So looks like we got a tool, came back.
Second part is, let's move Casey onto my team as a senior engineer. These are the building blocks of internal mobility. If this works, we could probably build a whole agent that does this. You will see that we are submitting some stuff, and it looks like it is done. Here is the problem. This was supposed to go through approval chains and workflows and business processes that we have set up, that our customers have set up inside of the tenant. None of that actually happened in this case. This is a big issue for that, okay? Let me show you another problem here.
This is kind of the other pattern that we see with our customers, which is I want to take the data inside of Workday, I want to extract it as like a snapshot, put it into a different system like Snowflake, which is what we see here. You can see we have got this Workday snapshot. Look at all these tables. We have got all these different tables here, all the different data inside of Workday. It is actually quite helpful to have this data in a lake. You can see here is our job requisition field. We should be able to ask similar questions. Let us try it again. Show me my open job reqs. We are getting 30 back, and what I am noticing here is this is actually account executive roles. This is the sales team. These are not the right roles for this agent.
This is actually a completely over-permissioned agent because remember, Maya is in the engineering department. Why is Maya seeing roles from sales, marketing, HR? This is actually not what we want. Just sort of a crazy example here, but we will go paste something else in. What is the CTO's salary? Just as an example. I mean, this should definitely not come back. We take a look, and here it is. It is coming back. I am sharing this because this is very real, right? When you extract the data out of Workday, put it into a data lake, all of the rules come off the data. These kind of things are not only plausible, this is in fact how this looks. Building agents against this is very risky.
Pointing an agent at the lake, for a lot of customers I talk to, it feels like the obvious next step for their data strategy. But what it is really doing is it is quietly turning a governed system like Workday into an ungoverned one. I do want to be clear, data lakes are awesome. They are actually great for lots of things. They are great for joining your other enterprise data with Workday data. They are great for building reports and analytics at scale, RaaS reports, things like that. But hopefully you saw in that example that you can export the data from Workday. You cannot export the rules that govern the data. Maybe you could try and rebuild them, but that really does not work long-term as a strategy. Until today, we have not really offered our customers a great answer to this dilemma.
But with that, let's see headless, now the lawful way. All right, we'll move back, and this time we're going to start on the data side. This is in Tableau, and we're going to take a look at this connection. You can see there's a client ID and secret. I know you guys probably don't quite follow exactly what this is. This is Maya, logging in as Maya, okay? And getting low-level access to Workday data through our Live Data Query product. Now, we're going to ask that same question, paste it in here. Show me my open job reqs, and what you're going to see is we're executing that query. Now, we got a different response. We've got 16 open job reqs, and I can tell by looking at these that these are actually scoped to engineering. So these are the right roles.
I said before, Maya needs to be able to do what Maya can see. So this is what Maya can see. Now, the other cool thing, and this is something I hear so much from customers, is, look, Workday has this unbelievable object graph that has been the secret to our success for so many years, right? It powers all of the 20 years of HR and finance success that we've had at Workday. But it's also been tough for our customers to get raw access to this. So what you're seeing here is a SQL query, that select statement there. And this is something our customers have been begging us for. Developers get raw, open access to live Workday data without making a copy that leaves the rules behind. All right, so let's move on to the MCP front here again.
And you can see this time, this is the official Workday MCP server, what we call our Agent-Ready Tools, now generally available. We'll go ahead and connect this up, and the first thing you notice is we got to log in, okay? Because this is going to be done as Maya. So we're going to sign in as Maya, and this is now connected, and then we'll go ahead and start some of these prompts again. Show me my open job reqs. And again, the identity piece is so critical to this to get right. And you can see that we're getting 16 open job reqs. These are all mapped to the engineering role. And again, if we go ahead and move Casey, we're going to see a better outcome in this case, which is not only are we going to find, hopefully, the right Casey- Which we did.
Casey, who's a software engineer in data platform. But what we also did is we started an approval process and workflow versus just executing the approval right away, which is what often happens in some of these other failure mode cases. Now, the other thing I just want to point out, because this is the really exciting piece, is check out how many tools we have. Look at this list of tools. About 400 tools we have. Read tools, write tools. I'm not going to read through all of these, but this is the surface of Workday. And by the way, we're constantly expanding this list. Now customers ask me, "How many tools? When do you get to done?" I honestly, I don't know we're ever going to get to done. There's always going to be new combinations of tools we're going to build for customers.
So incredibly exciting potential here through the headless enablement in Workday. Why does this matter? Headless used to force a choice between speed and control. Today, that choice is gone. Lots of great capabilities, but let's talk for a minute about what headless enables, what it enables for our customers. If you think about Workday, in some ways, Workday represents your company's institutional memory. I like to think of people and money data as the base of enterprise needs. Everything in the enterprise from a data perspective, it all derives from your people and your money. All of that context used to be locked away. Starting today, any agent that customers build anywhere can get to use that context and use it safely, as defined by our guardrails. What can this unlock? Lots of things.
Just a couple examples that come to mind. Field operations. Imagine an agent that is able to assemble a field crew, the perfect crew, the right experience, based on performance history, kind of pull that all together, do that without any data leakage. Really exciting. Maybe on the R&D side, maybe closer to my world, the agent could, let's say, model the financial cost of a team's big new bet. We want to go build something crazy. Do all of that right inside Jira, where the engineering team works. Maybe retail operations, where the agent can match staffing needs directly to point-of-sale traffic, balancing real-time availability with exact cost center performance. This context changes everything for our customers. When they have access to this headlessly, who knows what they're going to build, and who knows how much usage and demand this is going to drive?
We honestly just don't know because this is so early. My bet is, given how foundational authoritative people and money data is to enterprise workflows, the demand for this is going to be incredibly high, and we're already seeing signals of that now. To bring this all together on open by design, Workday is now headless. Agent-Ready Tools, GA, tons of momentum on this. The demand is just incredible for this, as I said. 500 tools, 400-ish, 500-ish. We're shipping about 130 of these a month, by the way. So the momentum is really accelerating in terms of our ability to churn these out and cover more surface area of the product.
By the way, I do want to mention, I think this was hopefully noticed in some of the news, but we have been partnering closely with Anthropic and some of the other AI labs on this MCP surface. They are our customers, and so there's not only the fact that we're shipping things, but that they're getting used by the likes of Anthropic is, I think, an important proof point. Workday Data Cloud. Also GA. Live Data Query, that capability I showed you before, SQL access to the Workday object graph. This amazing piece of technology that has propelled Workday for so many years. Now GA, Live direct SQL access to Workday data with the rules built in. The feedback from EA customers is, look, they want options.
As Gerrit kind of laid out those three options, they want to be able to choose the right path based on the right use case. Some paths are going to be headless, some paths are going to be sub-agents, some paths are going to go through the Sana Workspace and some of the unbelievable end product that you saw in Joel's demo. We win when our customers win, and that can happen on any path. Shipping tools is only part of the job. The other part is making these tools effortless for folks to use, because today, everyone is a developer, everyone's a builder. I like to reflect on the AI companies that I see. If you go to any of these big AI companies, you get to sign up for the product. You get to use the product.
You don't have to talk to a salesperson first. I think Workday has really thrived in that prior model, but that's not the future. The future is log in, sign up, get an account. When we do that right, it fuels a product-led flywheel that will drive pipeline on the sales-led side. So two key updates that I want to share on that front. Starting today, you can get a free account on developer.workday.com, and you can also get instant access to a developer tenant. That allows you to basically probe on the MCP surface. Fun little fact. We did the CIO keynote today where we shared this news. Biggest applause line of the day was actually this, because folks who've tried to get access to Workday tenants know it's been pretty friction-full, trying to get a tenant.
It's also been expensive, frankly, for Workday to provide these, because these systems are architected for pretty large scale. So we had to actually build and engineer a new kind of lightweight, small-sized developer tenant to create this product-led flywheel. But the net result is you can now go to developer.workday.com, sign up, get the tenant, build your headless app, promote it into the marketplace, and do this with whatever tools you want to use. So Cursor, Cloud Code, VS Code, whatever. We also have our Developer Agent, which is our prescriptive way of building agents. Think of it as like the guided path for building agents the way that we think they should be done. That product has been a rampant success. Since June, 125 customers have built 500 agents and 6,000 forms-based apps with Developer Agent.
In fact, I had one customer who built 25 agents over a weekend. This is a media customer. Totally insane velocity that we're seeing. What does this mean for financial levers? First, on the ecosystem front, we have tripled our developer community in two years. A lot of momentum here, but we are only just beginning. We believe that with this product-led motion, we're going to be able to 5x our developer ecosystem from where we are today. The second is marketplace expansion. There is still a huge need for content on the marketplace. That is, areas of the product that Workday first-party roadmap is not going to cover. We think there's a great opportunity for developers, whether they're affiliated with a service integrator or a major partner of ours or maybe unaffiliated, to build apps, to monetize those apps, and to take part in our ecosystem.
Thirdly, there's a need for a global talent pool, and we hear this a lot. Workday is obviously a global company, but we need help. Our customers need help implementing Workday, deploying custom AI, headless integrations, building advanced workflows. We need to make sure that there's a base of talent who understands Workday worldwide. These solutions that we're describing helps us create that global talent community. So product-led flywheel, developers, marketplace adoption, ultimately long-term pipeline. The final pillar and the crux of that commercial that you folks just saw, and this is really the secret sauce. This is the moat right here is lawful, right? The ability to map to our guardrails. Anybody can wrap an LLM. I think that's pretty clear at this point.
Nobody else has 20 years of HR and finance logic that is driving core business workflows for companies at the scale and import as Workday. It's just a fact. Agents with a capital L, or sorry, are lawful with a capital L when the reasoning and the execution run inside of the platform that governs it. In other words, the reasoning has to sit next to the system of record. I'm excited to share that Workday Agent Harness, which is what does this is now GA, the trusted runtime for enterprise AI. So walk into some of the details on this. Now, I want to be clear, this is unlike MCP, some of what you saw before. This is about keeping the execution and reasoning on Workday, sort of the token path, what we sometimes refer to as the better path.
This underpins every agent in Workday, by the way. So every demo that you've seen, all of it runs on top of this Agent Harness. What does it have? Memory, guardrails, sovereign routing. Sovereign governments I talk to a lot, they need to make sure that the AI is all within the country boundaries. Full visibility into reasoning and context, all of that in one place. What we're now doing is we're actually opening up this harness, from a tech perspective, and letting folks build directly against it, right? Making sure that these layers are all available directly to the customers who want it. Now, one point I want to make on context, just to try and pull out the differentiation here. Imagine that you're trying to find a successor at a company of about 40,000 people. Okay? 40,000 folks in the organization.
If you need to churn through, I don't know, 10 years of people and performance history, that's going to be about 1 GB of data, sensitive data, by the way. If you're using Claude or Gemini outside of Workday, you're going to have to painstakingly extract 1 GB of context across to some other harness, tool call by tool call, pulling 1 GB of data that probably shouldn't be leaving Workday, by the way, and send it over the public internet, by the way, right? With Workday Agent Harness, you don't need to do that. The context is ambient. Nothing moves because the system of record is right next to where the reasoning is happening. This is engineering physics, and this is part of Workday's structural mode. Now, why is lawful by default giving us a moat? Some other pieces here. LLMs lack business logic and permissioning.
Nobody else can match executing inside of Workday. We also have platform stickiness. This capability, these capabilities, Agent Harness, for example, it locks in our core HR and finance workloads while driving high-trust AI. It also solves the number one CIO fear we hear about, which is risk mitigation, and this idea and this concept of lawfulness. I think when folks start to see that they really perk up because intuitively they understand how much risk there is here. So a few takeaways for you. Open by design, new platform consumption revenue, backed by builders, this hybrid sales-led and product-led motion. Ultimately, our ambition here is to scale that ecosystem I referenced by 5x and fuel pipeline in the process, and lawful by default.
No one can match our HR and finance core that really derives a lot of the differentiation from the agents that run on this platform. Now, we are taking governance even further. I think you have all seen the C1 acquisition. Fun fact about this, we actually surveyed our customers about this, and granted it was a small-sized survey, but 70% of them told us that they want HCM and identity governance combined into one system. I will never forget, I was on the phone with a mining company, and the CIO told me, he was like, "Why are you making me buy a second system for this? This should be one system." I took that many months ago, and I was like, "Wow, there is something really here." Because identity is so central, as you can probably see to this lawful lawless piece.
It is all about identity at the end of the day. Workday, as Aneel has said, is the upstream provider. It is the source of truth from which identity derives. We are doing that today for humans, but as we look forward to AI, when we look at the growth of non-human identities, they are going from about 1 - 50 to 1 - 70-ish. In other words, 70 non-human identities for every single human identity. That is the projection over the years to come. This acquisition puts us in a really incredible spot to start to capture value around how we can govern that in conjunction with the rest of the org chart and the rest of the business processes that we provide on the HCM side. Do it all with one solution. So super excited.
Together, we are going to become this identity governance and security enforcement layer for not just people, but also for agents, and do that across the enterprise. So for customers, what this concretely means is no more accounts with keys to the entire enterprise universe. Permissions get right-sized to exactly what is needed, exactly when it is needed. One platform, one set of rails. That is the platform opportunity. I think with that, we are going to hand it over to Rob. All right. Thanks, everyone.
I've got one. Appreciate it. Thank you. Hi, everybody. Good afternoon. Great to be here. I am so excited to be able to take what you've just seen to market, and you can see from the keynote today, our customers are really, really excited. I'll walk you through our go-to-market motion, how it's evolving, and share a closer look at our customers' journey as they adopt our AI solutions. Let's just jump right in. As Zane and Gerrit shared, our total addressable market is substantial, spanning both new customer acquisition and expansion of our existing customer base. On the ground, I can tell you, almost all of my conversations are focused on AI, APIs, Workday Data Cloud, and how this connects with finance and HR.
To capitalize on the opportunity we're seeing, we're sharpening our go-to-market focus and investing across the following key growth areas: customer base, AEs, COE, medium enterprise, and global growth. Customer base expansion remains our single largest and fastest route to revenue. As Zane shared earlier, we are on track for AI to represent over 50% of our incremental customer base ARR for this year. We're committing substantial go-to-market resources to capitalize on the deep trust, platform stickiness we've built with existing customers.
This investment is broad-based within our customer base team, but it is strategically weighted towards larger customers. This is where we see the highest expansion potential and the most immediate, predictable pathway to cross-selling and growth, particularly as AI adoption accelerates. We're not just adding sellers, though. We're also adding specialized COEs to incubate, scale, and mature our growth products.
Our current areas of focus are on Sana, if you didn't get it, talent, platform, and spend across both organic builds and acquisitions. The COEs partner with account executives and work across the complete go-to-market motion, selling lifecycle, driving real execution, deal execution, strategy, enablement, evangelism, demand generation, marketing, and customer insights. Our forward deployed enterprise architects work closely with our customer pilots during the sales cycle to activate the agents and accelerate enterprise AI adoption.
Partners are key. We're empowering the GSIs such as PwC, Accenture, Deloitte, KPMG, to build differentiated IP and AI offerings that reinforce Workday's core specific to industry, personas, and geography. Over the past year, we have sharpened our focus on the medium enterprise segment. First, we focused our ME upmarket land team on the category of 500 - 3,500 employee segment. This is our sweet spot, and we're being more intentional about this.
As a result, we've seen improvement in win rates and an increase in net new deal sizes, nearly 20% year-over-year. We also simplified the packaging and implementation model and expanded Workday GO into more than two dozen markets. We're seeing good momentum as we continue our rollout and expansion, including strong win rates and faster deal cycles. In Q2, Workday GO volume increased more than five times over Q1. Within the ME customer base team, we've increased our capacity and aligned our teams with the COE sellers that I mentioned earlier. This is driving nice uptake of AI solutions within our ME segment. In fact, our customer expansion rates in ME are very similar to what we're seeing in LE. Switching gears to global growth. We're seeing strong performance out of certain markets like the U.K., Canada, Japan, and parts of Northern Europe, and ASEAN.
To sustain this global momentum, we're taking the following actions. We're opportunistically adding capacity in high-performing territories. We're opening up new markets targeted at international markets where we are scaling our value-added reseller program. We launched our VAR program in February of 2026, onboarding 20 unique partners, now reselling across 14 countries. Workday's AI is grounded in deep enterprise context, speeds up the customer's path to value by combining business logic, unified data, and strict governance. We are seeing strong uptake of AI across several areas. In total, 25% of our core customer base has adopted one or more SKUs, representing $600 million of ARR, and that's up from just over 10% of our customers a year ago. We're also seeing more repeat buys. This is a very strong signal of the value that our customers are achieving.
As Zane mentioned, the monetization we are seeing is driving a healthy balance, as you can see from this slide. The growing base of our customers builds a flywheel of value statements that we can leverage. In addition to building commercial momentum, we are really pleased with the adoption we're seeing across our newer initiatives. Over 50% of customers have signed our AI terms of service, and they are beginning to use their entitlements, and we see a clear path to 75% over the next few quarters. We have more than 125 customers and partners in our EA program for MCP, which is now in GA. We've already signed our first wave of customers on the Workday Data Cloud since our launch a couple of weeks ago, with a rapidly building pipeline.
We've also signed 18 of the most significant brands in the world in our Sana Enterprise Lighthouse program. These are companies that are deeply co-innovating with us over the next year and have agreed to serve as references. We forego one year of licensing, but we are thrilled with the partnership here and believe it'll lead to much larger long-term opportunities. While much of the $600 million of AI today is sold via traditional subscription license, a growing mix is coming from Flex Credits, a usage-based pricing model we announced last year at Rising. These are typically one-to-three-year commitments from customers that are tied to measurable outcomes, not token consumption. For example, when we sell Flex Credits tied towards a recruiting use case, it's based on the number of applicants. We are very early in the Flex Credit model, but momentum is building rapidly.
We have nearly 300 customers that have moved onto Flex Credits, including nearly 200 customers we closed in Q2 alone. There are a couple of things that's happening in the background that should drive an accelerated uptake of Flex Credits as we move forward. One, our first-party agents continue to mature, and we are accelerating the amount of innovation we are bringing to market. You could see it here today. Because of the Flex Credit model, which is fungible across our agentic portfolio, the more solutions we bring to market, the more measurable value our customers will get, and the more appealing the Flex Credit model becomes. Two, we are opening up our MCP and Data Cloud programs into GA right now, which should also lead to increased momentum as customers leverage the power of the platform. Let's take a look at two very distinct customers.
I think the one might be the same as Gabe mentioned. Putting Workday AI to work to reinvent how their businesses hire, operate, and scale. Our first story is about a healthcare company with nearly 70,000 employees. They turned to Workday to optimize their data in a runway for AI, enabling them to maximize resources and focus their dollars where it matters most, the patient. In FY 2025, they adopted HiredScore, which enabled them to develop the AI governance strategy in line with Workday's responsible AI model. That foundation paved the way in FY 2027 to add Flex Credits to give them capacity to roll out solutions, including Self-Service Agent. Ultimately, AI adoption across this journey has driven a 50% uplift in ARR. Our second story is a large manufacturer/mining company in Asia Pacific with over 60,000 employees.
Initially, they came to Workday to solve challenges with manual recruiting tasks, long hiring cycles, excessive administration workloads for their managers. What started as a single HR platform deal has grown into a strategic partnership, a cycle of investing, measuring the result, and then expanding. In FY 2026, they adopted HiredScore to reduce the burden of manual tasks on talent admins. Nine months later, they closed a deal for Paradox to layer conversational AI and candidate nurture on top of it. Today, we are seeing continued investment outside of talent with Evisort, Extend AI, and Sana. The payoff? They've automated scheduling, improved candidate communication, reduced hiring time, cutting the time of managers spending on recruitment in half, and their AI journey has driven a 40% uplift in ARR. This one I'm really excited about.
In the last few minutes, I want to talk and quickly touch on the opportunity that I, and I know the teams are very excited about. The CIO is an emerging distinct buying center alongside HR and finance. This dynamic is one of the many reasons I'm excited about today's announcement. AI is pulling them into the Workday conversation in ways we haven't seen before, and this represents a huge opportunity for Workday. To capture it, we'll be focusing on the CIO persona with our innovation offerings like Workday Data Cloud, Workday Extend, and our AI agent use cases. Since we announced Workday Data Cloud last year, we've moved from concept to GA in less than 12 months. We've seen tremendous interest with more than 1,000 enterprise customers of ours that are leveraging Snowflake and Databricks for data warehousing and have already closed 20 deals with rapidly growing pipelines.
Building on this momentum, we're also expanding our partnership with Google Cloud and AWS to deepen integrations and broaden our reach across cloud ecosystems. The deal economics are already proving out. Platform adoption is driving meaningful ACV expansion opportunities for Workday. We are seeing growing interest in Workday Extend AI to build both traditional and agentic applications on Workday, and there are two different types of opportunities here. Convert the more than 1,000 customers from traditional Workday Extend towards Workday Extend AI.
This drives a roughly 20% increase on average to what they were paying with the traditional Workday Extend license. Two, drive first-time platform customers, and this is the biggest driver of growth with Workday Extend AI. It's also worth noting that for customers to leverage our Agent-Ready Tools and Workday MCP server, they need to be licensed on Workday Extend AI as the first step, and this is helping drive momentum.
We are just getting started here. Nearly every customer I meet with is excited about leveraging our platform and MCP for certain use cases. Our MCP program is in GA, and I believe this creates a tailwind moving customers onto Extend AI and Flex Credits. Sana Enterprise is a central AI orchestration engine across the enterprise. You all heard exciting announcements this morning about C1.ai, which opens up a much broader opportunity with the CIO around identity governance, a top priority for many customers that we believe only accelerates with the proliferation of AI. It's an incredible opportunity for us to leverage our trusted position within the enterprise and C1's innovative technology to drive adoption. To bring it all together, the runway ahead of us is vast. We're excited about it. Our teams are excited about it. You can feel the customer excitement.
We're excited about how this go-to-market motion will evolve over this year, and we're confident that this will position us to drive rapid customer AI adoption and capitalize on a TAM that's growing both the land and expand opportunities. Thank you for your time today. With that, I want to invite back the leadership team, and we're happy to take questions. Thank you. They're going to bring a chair out for you. Thank you. Thank you. Thank you. Oh, you guys now.
Thank you.
Well done. Well done.
Chairs, finally. Okay, you have the first pick.
I'm going on the end.
Where do you want this, Team?
I was told to sit here. Oh, no, I can't come. No.
Okay.
Sitting down is nice.
That's a nice. Thank you.
We are going to go around the room. I know there's a lot of people, and we're not going to get to everybody, so I apologize in advance, but why don't we go, Aneel, on your side? We'll start with Alex. If you could just say your name and then make sure they can see you.
Yeah, it's really bright.
Alex.
The button turns it on. I work in tech.
Alex Zukin. Thank you guys for hosting. Alex Zukin here with Wolfe Research. A really inspiring presentation, at all levels, the tech, the interface, the modules, the TAM expansion opportunities. I wanted, Aneel, maybe to ask you, it feels like there's kind of a big debate about headless becoming the new interface layer for interfacing with enterprise workflows as much as applications or software agents. What do you anticipate those economics to look like? Some of your other kind of compatriots are leaning heavily in, I think as you guys are, on the headless strategy, and consequently, they think they're able to derive significantly higher unit economics from that strategy. Then there's some others that are leaning very hard in the other direction.
Yeah.
How do you expect this balance of trade to result when you're interfacing with customers? Then maybe Zane, for you, I know we're not doing midterm or long-term targets here, but maybe just the conviction given the amount of innovation you've demonstrated in maybe the sustainability of double-digit growth as an aggregate kind of concept.
Sure.
I guess at the highest level, I don't think anybody really knows how it's going to play out between headless and your own front end. Now, I think a lot of our competitors and peers don't have anything like Sana, so they default to headless because they don't have their own prompt-based, AI-driven user experience the way we do. I'd say the other piece is we are an HR system. So we do naturally touch every employee in the company, not just a group of employees, maybe in sales or marketing or services. As HR, you touch every employee in the company. So I do think we have a different strategy. Now, if people want to do headless, they work through our APIs, and my expectation is we'll be margin neutral. The beauty of API is it requires a lot less sales cost.
I'd always like people to use our user experience, but if they don't, our guardrails that are exposed through our APIs are super valuable. That's the real value of the underlying system of business processes, of security model, of org chart. I think we win either way. I think so far what we're seeing is customers really like our strategy where we have a user experience that's best in class, and that's only because of Sana.
Just to add, today we didn't announce APIs in Claude, we announced Sana in Claude, right? This is an AI to AI integration, which means the reasoning, the intelligence happens on the Sana side, right? I think the interface is a bit of a blatant relevance. The real question is—
Yep.
—where is the reasoning?
Where does the reasoning happen?
I would just add, getting beyond FY 2028, we intentionally wanted to highlight for all of you the incredible work that this team is doing on these agents and the receptivity and adoption that we are seeing from the customer base. As Aneel put it best earlier today, our intention is to focus on growth and to drive growth, and we are excited by this consumption model. I would end it there, but our intention is to lean in and continue to grow the business. That is the intent. As Aneel said, we will see how things play out. There is a lot of interest and a lot of adoption to come, but we feel like we have got a tremendous portfolio to drive that.
Hi, this is John DiFucci from Guggenheim. Gerrit, you said that other than coding, the AI breakthrough has not really happened, or not yet anyway. It took a long time for the large enterprise to really adopt internet-based applications. Aneel and Dave saw that coming and got into it before it happened and actually was ready for it when it came. I guess, how do you think this market evolves for the use of AI by businesses, and how is Workday positioned differently, and this is really, I think, for Gerrit and Aneel, differently for this major technology paradigm shift than the last one?
You want to start? Should I start?
You are smarter. You start.
Okay. I think that was the smart way of-
If you haven't figured out, he is smarter.
The smart way of handling that question, actually. The reason is that for AI transformation to really happen, companies need two things, right? One is they need to actually understand how to transform their existing work, right? The success in AI is not in applying it to the work you already do. The reason why we see so much success in coding is because code, to a large degree, is language. It's a verifiable problem. There is high proficiency in the models, per se. So there's not a lot of domain specificity to it. There is so much code on the public web that you can train models on, and it's mostly driven from a very simple business process, right? You write code and you submit it, and you should review it maybe in between, and then you deploy to production. So it's a very simple workflow.
Now, if you take any other enterprise workflow, right, which strings together multiple functions. For instance, finance and HR. You have hundreds of steps, thousands of variations, different people interacting with each other. One key part, which is why we announced Work Intelligence today, is actually getting to comprehension and understanding about how does work actually map to the language of AI models, which is tasks. Models fundamentally don't think in jobs and job profiles. That's anthropomorphizing that space. They have task proficiency. They can do one thing really well and maybe another thing. But the thing in the middle, they can't do really well, right? So you have to extract all of the things they can do from the things they can't do, and maybe also the things they shouldn't do. Then you have to project this on a new workflow configuration inside of an enterprise.
That is the fundamental step for at-scale AI adoption, and that is what makes Workday so strategic, right? Because fundamentally, you talk about an HR people and organization transformation first and foremost, right? Usually, as saying, keep saying, technology comes last, right? First, you actually have to change the business and the process. That is why Work Intelligence is such a big stepping stone. Secondly, this is why C1 for us is such a big part of our strategy, is that even if you have an understanding of where you want to apply AI to, now the question becomes, how do you do this in a secure, reliable, and lawful way? The problem is that none of the recipes that you have in classical ERP, let us just say how we thought about ERP, enterprise resource managing, right?
The bills of material and the supply chain and blah, blah, right? Everything that was built for that does not work for AI. The unit of work is different. The cost model is different. You need to have now real-time economics based on real-time spend against tokens because it becomes such a big part of your operating expenses. You need to think about security, right? How do you do this and stay compliant? How do you do this and stay safe? So you have to fundamentally rethink what the platform an enterprise runs on, which used to be ERP, actually does. This is what Agent System of Record C1, right? This is why we are bringing all of these assets together because we think with that work transformation, also the business applications are going to reconfigure themselves around agentic identity, agentic work, and agent to human collaboration.
Yeah, I guess the only thing I would add to that is just look where enterprises have adopted AI and specific to our world, kind of on the edges so far, recruiting. Not super high risk, and it is a pretty contained environment. Same thing with what we are doing with Workday Contract Intelligence. This next wave of agents actually go deep into the bowels of an organization, and it requires real transformation. The best place to start is at the top, which is why Sana and Self-Service Agent or the Concierge product is so important. Then you start drilling into the middle. The financial audit agent, which is my favorite agent because it is going to automate the financial audit process, will be really powerful. That will require a lot of workflow changes.
That makes a ton of sense. My understanding is within Workday, you are all using the Sana interface.
Yes. Excessively. Same worries sometimes.
Yeah. We worry about the token usage.
But I think most of us here probably are Workday customers. I know we are. I think, like you said, it starts from the top. Once we get that, when will we get that?
You have it now.
We have it now?
Just have to-- What do they have to do?
Today. No, well, you have to turn it on, right?
Okay
It's now the standard home screen or the home experience for Workday is Sana.
Okay, great.
Yeah.
The example you showed, Gerrit, just taking time off.
Yeah.
Yeah.
Just call your IT team and tell them to turn it on.
Thank you. Thank you, guys.
Tell them the button turns it on. It works all the time.
Just click the button.
Over here. Kirk Materne with Evercore ISI. Thanks for the day. Gerrit, Joel, Gabe, you guys gave a lot of great, really intuitive examples of how agentic sort of integrating into the platform. The question's for Rob, which is, you guys have a lot of customers that are using Workday, that are paying you good dollars to use the system as is. How do you make sure that you transition the value created in a lot of these newer products into incremental revenue and those customers? Some of them are used to buying per seat. To Aneel's point, they have to rethink sort of the processes around this.
Yeah.
What is the opportunity over the next 12 months to make. You just have to sort of not give away, but seed the products in them so they start to see it first, and then you monetize after the fact? It seems like there's a lot of avenues to pursue. I'm just curious where you think the lower-hanging fruit is and—
Yeah
—how you get there.
It's a great question. I would say, the appetite for our customers has evolved rapidly over the last six, seven months to understand how, first of all, the AI SKUs fit into it, how we build a recruiting end-to-end solution for them and bring that to market and determine the value. It's really value-driven, all of these equations. We have to change the sales cycle that was traditional seat-based into value-based selling type, and really showcase customers how they get the value. The big thing for existing customers is, this is not like you sold software previously, where you sell other values X, and you're going to gain it in 24 months and just wait and see, and it's going to really feel good. This is about showcasing in the moment, like getting these agents activated, deployed. We've taken an extremely thoughtful approach, I would say.
When we brought Self-Service Agent to market, we worked with 10 of the largest companies in the world, bringing the most complex self-service skills into the market and bringing them live on a day to actually a week-to-week basis. As we get that hardened, now we have an opportunity to scale it and scale it. I think that's how we will drive it across all of the customer base. With new opportunities, we change the sales cycle because the value equation now becomes agentic as well, and we build a new value equation into that customer base. I would say the seat-based pricing is the baseline, and everything in the future is going to be value-driven. Managing the cost model for the customers and on behalf of the customer I think is really critical to be successful.
Hi. Tal Liani from Bank of America here. I have two questions. AI is growing. Also, you're focusing on medium-sized enterprises, but still subscription is slowing down. I'm talking about the next year, the next two years. Can you define the strategic challenge? Why is growth decelerating, subscription growth decelerating? What's happening in the market that we don't see stabilization or acceleration before AI? The second question is about AI. I see a lot of similarities between your strategy, Salesforce strategy, ServiceNow strategy. Customers will have to deploy. If they deploy all of you, they'll have to deploy multiple identity solutions because each one of you has a solution, multiple interfaces because you have Sana, they have Slack and others. How is it going to play out?
Are they going to deploy three identity solutions and three interfaces solution, or are they going to deploy an external one by someone like Microsoft, maybe, to harmonize everything? I am trying to understand from which direction this market is going to evolve. Thanks.
I will answer the last part first, having spent a long time in identity. HR is the system of identity. We built the Agent System of Record to couple with our HR system. There is no question, no one will debate you that HR is the source of identity for all enterprises. Where else do you get your source of identity? ServiceNow doesn't have it. Salesforce has it for salespeople and the people that are in their system. But the one thing about every company is that everybody in the company is an employee, and that is where you get the identity. And we were architected for that. It is why I don't think there might be multiple identity systems in a company.
There will only be one source of truth, and that will always be the HR system for employees and the Agent System of Record for agents, and then the governance of what those employees and agents can do, which is what C1 does. I don't really think it is that complicated. I just think we have the winning hand because we are that core system of record for identity.
As a matter of fact, today customers have multiple identity systems.
Yep.
The HCM system sits upstream from that. The multiplicity in identity system today is the problem they are trying to solve for.
That is where the governance comes in, which is why C1 is so powerful. Because if you have multiple identity systems, one HR system for identity, multiple identity systems that are authenticating and giving access to applications, well, then how is that governed in a centralized way? That is where we come in. We basically think about Workday as surrounding the IDPs. What was the other part of the question?
To understand the growth deceleration, the source.
Without talking about acceleration, I will leave those things to Zane. I will say, the thing about medium enterprise is it is a massive part of the market. It has always been important to Workday. The biggest issue we have had historically is that the cost of our deployments have been more expensive than folks like UKG, as an example. Cost of software, close to one-to-one, but our deployments might have been two to three times the expense of a UKG, partially because they subsidized the deployments and partially because we built a highly configurable system for large-scale enterprises, and going through those choices were hard for medium enterprise customers.
With the deployment agent that Gerrit and team have come out with, we dramatically reduced the cost of the deployment. Completely changed it. We have already seen it in the win rates as people start using the deployment agent. I actually think what's going to happen is you're going to see UKG customers leave UKG and come to Workday because UKG doesn't have a strategy for agentic technologies. If you're a company that wants to embrace what's happening in AI, you get a really fast conversion over to Workday, and then it opens up all the AI capabilities that we have. I think you're going to see that happen across the board.
Yeah, I'd just add to that, the model's changing, and what you've heard today is as we move more towards agentic, towards consumption, and that value-driven opportunity, it takes time, and we're doing things like lighthouse programs, and we're really driving future growth and trying to get that flywheel going. We've been very transparent around renewal cycles, M&A, lapping year-over-year and things like that, as well as international, which at times has been a huge propellant for accelerated growth, and at other times it hasn't. If you take the noise out, I'd say as we continue to move this model more towards what you saw today, we're very excited about that medium and long-term opportunity. But you've got to grow into that, and customers have to adopt that.
Mark Moerdler, Bernstein Research. In the last year, you've come really far on the technology, and the capabilities you've shown today are very impressive compared to what existed before. I'd argue you've even done that on the vision side with the explanation of lawless versus lawful, and try and explain the value. I would like you to answer the question, and I know, Aneel, you're going to have a thought, but I think others might, on the question everyone asks us. That is, when you expose your data, your system of record, your semantic knowledge, through to someone else's, a Claude or someone else, a ChatGPT's front end, how do you assure that you maintain the customer relationship and the economics on a go-forward basis rather than the front end being the capturer of that upside and that economics?
Well, I think simply for those tools to work the way you described, they have to go through our APIs. If you just go directly against the data, you lose all the context. You have no idea of the business process, what guardrails are built around it. We have that example of query the database for the top five sales reps. You're the Northeast VP. Well, without knowledge of the org chart, without knowledge of what you can see as that Northeast VP, you'll get top five sales reps in the world. That's a very basic example, and Gabe came up with that example. But to describe that situation, that's the headless situation. They have to go through our APIs, and those are fine with us because we get paid that way, too.
I just don't see HR and finance customers going around APIs and taking the risk of screwing up a payroll or screwing up a financial audit because they didn't have the context of the underlying business process. I've yet to meet a customer that wants to do that.
One other thing I'd add to this that I think is maybe at risk of getting missed here is, as Aneel said, yes, they're going through our APIs. But the laws and the rules and the governance that those APIs provide are actually being provided through the core HR system. So one does not exist without the other. The value of the APIs exist because—
Yep.
—the core system is providing the tenant configuration, a business process configuration, all of the approval workflows. So, it's those two things working together that allow us to maintain that direct customer relationship. Look, if they're coming in through our AI APIs and our MCP surface, as we've said, we're going to monetize that as well. But I think over time, there's also another dimension of this, which is we look at a progressive upgrade possibility where folks can start off on an MCP surface in a true headless fashion, maybe progress towards using sub-agents, à la what we've talked about with the Sana and Claude piece, then ultimately progress towards using the full Sana AI workspace.
We think that's going to be a use case by use case journey, but we do see the ability for customers to sort of migrate across different paths, which gives us optionality to make sure that we are controlling the customer relationship.
Okay. We're going to do just—
Yeah.
—sorry, two more questions. I know we've got Siti over there, and then we've got Dan over here, and then we're going to break for happy hour after this, so go ahead, Siti.
Siti Panigrahi from Mizuho. There are a lot of innovation today. It's great to see. As you know, most of the time it's a customer-driven innovation that's an enterprise application we have seen. But seems like what you're today showing is a lot of new innovation you are trying to push to the customer, like how the work is changing. How are you going to sell this to customer? How are you going to convince them to change the way they work and so that what you are selling and so that they leverage your product? It's a new innovation that you want to change the way they work.
But I think I would categorize it differently. Actually, in the opposite. It is the customers, it is the market coming to us. Every company on this planet right now asks the questions, how do I transform with AI? How do I become an AI-led organization? As Aneel said earlier, based on the many conversation he is having constantly, they are turning to Workday as the source for two of those elements. As an HCM system, how do I do this structurally? As a software vendor, what capabilities can you give me to drive this in HR and finance? So it is incredibly customer-centric because you tell me the organization who is not trying to figure that out.
Yeah.
What we do is that we are not then you are just taking this trivially, but we actually use that opportunity of AI innovation and rethink at the core every process in HR and finance and in IT to completely transform it through the use of frontier AI models. This is what we have seen. It is reimagined based on our domain understanding, but based on the request of the market of transforming with AI.
Yeah. I just add to that, we listen to the customers. We have 20 agents. Maybe some of them are not the right agents, but we will keep iterating and learn from that process. It is getting back to that startup mentality in terms of innovating with our customers. There is nothing on the roadmap that a customer has not asked for.
Dan Jester, Bank of Montreal. First of all, I have used the self-service agent. It is great. So thank you so much.
Yeah. You guys are the best. You were the first ones to really drive it. Thank you.
Thank you for going through the school of hard knocks.
Yeah.
I wanted to ask the question, and Aneel, you had talked about this. The first generation of agents are a little bit in a sandbox, like they are defined. The next generation of agents are going to actually require business process change by your customers. Can you just give us a sense how your customers are actually thinking about that? What is the role of the ecosystem and partners driving that change? How much can you do and help versus how much it really makes you have to wait until your customers are ready? Thank you.
Yeah. I will push the ecosystem question to Rob, but I would love Joel to answer the question on I think actually the self-service agent and how you deployed it, Joel was front and center with that. That is actually how we are learning from customers, and how it is not an edge case like recruiting, but actually something that is central to how an organization works.
Yeah, if you think about something as this sort of this Copilot experiences. At best, you will make someone 10%, 20% more productive. As we transition to this more autopilot-esque experience, you're really re-engineering the entire workflow to be handled by these agents. We have a few different ways in which we support that. First is our first-party agents. You can think about there as we've done the sort of full last mile. You can just deploy these agents. They're turnkey. We've sold them soup to nuts. The second is the experience that we have in the agent builder, where we offer you templates, but you can build any custom agents. How I see this play out is for customers, they'll have hundreds of thousands of agents ultimately. The work of the employees will be doing the sort of meta work of creating these agents.
You will be doing less of actually carrying out the work and more building the agents that do the work. As that role shifts, your role shifts towards redesigning the processes to be handled by these agents and be increasingly handled by these agents. First generation of a lot of these systems, I think, was a lot sort of 10% productivity gains. I think when you start seeing these orchestras of agents handling the work with increased autonomy, the workflows will need to change entirely. First, we want to build the best sort of pre-built experiences for that are turnkey, you can just turn on and run with, but also enable you to build thousands of that on the Sana agent builder product.
Just on the ecosystem, just like we've had it change dramatically in the last 12 months, I think all of the ecosystem players, whether you GSIs, whether you are local providers, whether you're BPO providers, your business is changing, and you better change fast. If you look at implementations, typical way, you don't see configuration as an important part anymore. What's left? The business model innovation is what's left. I think we will be, in the next couple of years, going through a major shift in business innovation, and new business models will evolve, and we feel really good about where we sit.
Okay. I think we are going to end it there. Thank you guys so much. That does conclude the Analyst Day for 2026.
Thank you.
Thank you.
Thank you, folks. Thank you.
Appreciate you guys all coming out.
Thank you, guys.