Elastic N.V. (ESTC) Earnings Call Transcript & Summary

Great. Hey, everybody. Thanks for joining us. My name is Koji Ikeda. I'm one of the software analysts here on the BofA enterprise software team. Absolutely thrilled to have Elastic, Ash Kulkarni, CEO, here with us today. Thank you so much for doing this. Super, super appreciate it.

Thank you for hosting me. And thank you, everyone.

I think I was talking with Ash before this conference -- or before this session, and I think I do have the pleasure of hosting your first fireside chat, you solo. So thank you.

An Elastic guest.

Thank you. Thank you. Thank you very much for doing this. So I guess for people in the room that may not be familiar with the Elastic story, could you just give maybe a brief few minutes on yourself, your background? And then from a big-picture perspective, what is Elastic? What are you trying to do about there? What is the opportunity that you're disrupting?

So Ash Kulkarni, I've been with Elastic now for about 1.5 years, pretty much spent my entire career in data and analytics and security in the last 6-or-so years. My focus was significantly on security as well, but I'm -- at this point, I consider myself to be a native of San Francisco, the city proper. I know that's something that you and I share in common. But I grew up in India, and now I have lived in the Bay Area for the last 24-or-so years. My career started as an engineer, so I'm an engineer by training and spent time at Sun Microsystems, Informatica, Akamai, McAfee and then came to Elastic. And I was both a buyer and a user of Elastic before I came to Elastic, so I had a pretty keen understanding of what the technology was capable of doing. I had used it in 2 different contexts: one is to build an Observability solution, one is to build a Security solution. So I had a very good understanding of the flexibility, the performance, how the platform could morph to do things that were really incredible. And so I came to Elastic as the Chief Product Officer responsible for products and engineering, and then earlier this year assumed the role of CEO. So it's been a wonderful journey and now really enjoying what I do here at Elastic.

Great, great. Thank you so much, Ash. So we're very lucky. You guys reported last week, so I wanted to definitely ask you this question right upfront. We thought you had a great quarter. I guess, from your view, what are some of the key highlights from your fiscal fourth quarter results?

Yes. So Elastic is -- for those of you who are not that familiar with the company, we are effectively a data analytics platform, but what's different about Elastic is we are really powered by search. Search is the core differentiator. It's the secret sauce. And the use cases that we typically get used for are Security, Observability and Enterprise Search. So Observability is a little over 40% of our business. Security is around 25% of our business, and the remaining is Enterprise Search and the sort of the long tail of other custom search use cases. And our focus for some time now has been on cloud, and Q4 was fantastic because of the progress that we saw on all these fronts. So in terms of cloud growth, constant currency, grew by about 72%. In terms of traction in cloud, we saw progress with all 3 major cloud hyperscalers. We released new capabilities in terms of the product and the platform itself. And Q4 really capped out our year very, very nicely and growing at the rate at which we grew, and we saw progress, both in our net expansion rate. Our net expansion rate in cloud is now over 140%, and all of that led to us getting even more confidence in the progress and traction of Elastic Cloud first and foremost. And we talked about the fact that we expect to reach the 50% mark where cloud will account for more than 50% of our overall revenue by Q4 of fiscal '24, and that's sooner than we had previously anticipated. And we also gave guidance, a multiyear guidance on expecting to cross the $2 billion mark in the next 3 years, so in FY '25. So very exciting about the progress that we are seeing on all the 3 major solution areas, our progress in cloud and serve as a nice way to end the year.

Got it. Thank you. Thank you so much. I wanted to dig into that $2 billion target a little bit from here. The $2 billion target by fiscal 2025, big number, implies 30%-plus growth CAGR from here. A lot of the fears right now in the investment community with enterprise software as a whole is can numbers potentially get cut or decelerate more than what people are expecting out there over the next few years? And you guys come out, guns blazing, $2 billion target, very, very bullish target. What are you seeing out there specifically that gave you that confidence to put that target out there?

Yes. So I don't know if I'd describe it as a very, very bullish target. But the way I think about it is, first and foremost, we track the macroeconomic situation very, very carefully, as you can imagine. The cloud business is a consumption business, and so we track consumption very carefully. And I personally spent time meeting with customers constantly. I was in Europe at least twice last quarter, spoke with customers and wanted to get a first-hand sense of how are they thinking about prioritization, spend prioritization and so on. This is where the fact that we are not a general-purpose database really is important to understand. So if you think about the use cases that customers tend to use us for, it's Security, Observability for Tier 1 applications; App Search and Enterprise Search where we are that search bar that's powering your -- say, your e-commerce application. Security and Observability are not use cases that you tend to prioritize down. Like if anything, these are the things that our customers are prioritizing about everything else. Because given the current macroeconomic environment, there's also a geopolitical environment that you need to consider. And you can defer an ERP upgrade by a couple of years. You could put off some IT project by a year or 2. The last thing you want to do is cut spend on your sim because you missed something in your data, and you hit the front page of the journal. That is not where you want to optimize, and we keep getting these signals over and over again in terms of -- even some of the customers I talked about in the last earnings call, customers like Swift. And you think about the use cases that they use us for as billions of transactions happen in the Swift network, they cannot -- I mean, this is mission-critical for them if there's ever such a term as mission-critical. So there's a lot of resilience in the use cases themselves. In terms of cloud, we've talked about the fact that our cloud has -- the mix shift has accelerated. That mix shift was changed by about 6 percentage points in FY '21. It changed by 8 percentage points in FY '22. So it now -- cloud now accounts for 37% of our overall business, and so that trajectory is moving in the right direction. Our net expansion rate in cloud is over 140%, like we talked about. It's a much more friction-free way for customers to onboard, start to use us for more and more use cases. And we are also seeing that as people think about where they want to optimize, there are conversations about how do you try and think about spend in such a way that you can use one platform for multiple things. And the fact that we tend to be consumption-driven, so you can start to use us slowly and grow, and the fact that when compared to some of the incumbents, our solution tends to be cheaper is a positive. And we are seeing that, not just in terms of greenfield, but also displacement opportunities where we are starting in one place and then expanding to take on more and more use cases. So all of that gave us confidence. And, look, our overall net expansion rate is 130% roughly there. And you look at the CAGR of the $2 billion number over a 3-year period, that's a 32% CAGR. So it's -- the natural expansion in the business is already such that it's not a target that worries us.

Got it. So when we break down that $2 billion target a little bit, and you mentioned exiting fiscal '24, 50% -- more than 50% coming from cloud. So that does imply for this $2 billion target more than $1 billion of breadth coming from cloud. And you mentioned right now in your comments about the frictionless being a positive way to adopt the cloud. So I was wondering if you could just dig in on why the frictionless part from a customer's lens is so important or will drive that adoption of the cloud in the future?

So it's something that we've already seen, right, which is why the cloud net expansion rate, like we said, has been growing over the last several quarters. When you're using Elastic in self-managed, although the functionality might be similar, the fact is that you need to worry about setting up the Elastic installation. So there is a non-zero effort in terms of management of that cluster and setting it up to perform well. The second thing that happens is, as you grow and data constantly grows, which is why even in our self-managed business, there is natural expansion that happens at those workloads. As data grows, you need to buy more capacity, hardware capacity, first and foremost. So if you're using your own data centers, there is a natural friction that makes it harder for a customer to scale the way they would want to. Even if they're using public cloud and using Elastic software in self-managed mode on public cloud, they still need to come to Elastic for new licenses as their usage is growing. And then all of those become effectively friction points for them. In the cloud, they're just purchasing a certain amount of capacity. But what we typically see is whatever be the initial capacity that they're purchasing, it's credits. In cloud, we sell credits. It's not like they're purchasing physical capacity or anything. When they buy those credits, they typically tend to consume those credits faster than they expected. And we see that all the time. So customers will get started. They don't need to worry about the management or monitoring of the environment. It's all done for them. It's our global SRE team that's doing all of that work, and they just start ingesting data. It's easy for them to use us for one use case and then expand to the second and then the third use case, and they never need to have a conversation with us because we don't sell our capabilities as separate SKUs. So you don't come and buy Elastic Security from us or Elastic Observability from us. You just come and buy Elastic. That's it. And you can step up to a higher tier, and different tiers have increasing functionality in all of our use cases, but that model makes it incredibly viral. They start using us. They expand. They bring in more data. They realize they might have started with log analytics and they realized that, hey, I can start using Elastic for APM. And we see that all the time, and they'll start instrumenting a few applications. And that -- those APM traces will start coming into Elastic. They'll get more value out of it. Over time, we displaced their existing APM vendor. And that just becomes a much nicer model, and it's much more friction-free. And we see that over and over again, and that's why we are so laser-focused on cloud.

That makes a lot of sense. And you mentioned cloud, I understand pricing model is usage. I've been on the web page fiddled with the different drop-downs, got to some sort of price, I was like, okay, I understand that. But maybe for those on the webcast or in the room that aren't familiar with the self-managed side, how does that pricing work?

Yes. Even for self-managed, we sell the notion of ERU. So it's this idea that you're buying effectively -- it's still a consumption model, but we don't have a way to meter it, right? There's no automated metering. So effectively, you're saying that you want to be able to use the equivalent of so much processing capacity and so much RAM and so much storage effectively. And then if you believe you're going to exceed that, you need to come to us and buy more. In the cloud, what's different is it's being self-monitored. In real time, you're seeing what your usage looks like. And as your usage grows, your credits just start depleting faster. If you deplete all your credits, you just start paying us effectively on a monthly basis. And well before that, our sales team will go to you and say, "Hey, you're burning through your credits faster than you might have anticipated. Do you want to increase your capacity, make a commitment -- make an annual commitment sooner than your 1-year mark?" So in that model, the model is very similar to what you see from the hyperscalers, the cloud hyperscalers. And even on self-managed, it is tied to consumption in some ways, but it is not as immediate, if you will.

So the cloud just makes it more frictionless from both aspects.

Exactly, exactly.

Wow. Okay, okay, okay. That makes a lot of sense.

It's a simple story.

I wanted to ask you a question on the stickiness of the product. We believe, definitely, it's mission-critical. You guys are powering applications at the core. You're doing Security. You're doing Observability. But talk to me a little bit about what it would take for a customer to say -- one day to wake up and say, "You know what? I want to take this out." What would it take from the customer's perspective to actually rip out Elastic? How difficult would that be?

I mean, it's -- if you are a core SIEM implementation, you're being used for threat hunting. You're being used for analyzing all the many tens, in some cases, hundreds of terabytes of data, that's coming in every day from all of your network logs, from all of your endpoint devices, sometimes from your physical security sensors, like taking that out is -- it's not -- takes years. It's not a matter of days or months. It's a very, very, very intensive process. And you would never want to do that, unless you found something that does even better or gives you a disproportionate cost benefit. And in our case, we tend to be really, really competitive from both a functional standpoint and from a total cost of ownership standpoint. And that's what we hear -- that's one of the big reasons why we are seeing the success that we are seeing in Security in terms of not just growing in greenfield opportunities but also displacing incumbents. Security used to be about 20% of our business in the past, and now it's on 25% of our business. And that growth is coming because we are able to not only demonstrate the functional superiority but also show that you can leverage Elastic and get a better total cost of ownership than what the incumbents might be able to offer. And that then becomes the reason why somebody is willing to make the -- take the difficult journey and start to sort of use us for more and more of that use case. And over time, we will displace an incumbent.

Got it. I wanted to ask you a question kind of on the core functionality, the technology of Elastic. I'm fairly new to DevOps. I've been covering the category for about a year. But over the process, I spoke with a lot of developers along the way. And I don't use the term lightly, but the developers say Elastic is best in class at what they do. The ELK Stack, it's difficult to replicate. What is it about the technology that is so differentiated out there?

It is the most scalable, most high-performance search platform out there, like period. There's nothing that -- and this is where having been a user and a buyer I had a lot of opportunities to sort of inspect it in every way I could and give me a lot of confidence in how the technology has evolved. Look, we -- Elasticsearch has been one of the top 10 most successful open search -- open-source projects ever, right? So it is massively widely deployed. There is a huge community of users that has expertise developing on top of Elasticsearch. And oftentimes, they have downloaded the free version, and they're using it out there. So there is a massive base out there. And as we move to the cloud, as we make it easier for people to start using our capabilities in the cloud becomes a natural opportunity for us to basically say, "Hey, this thing that you love, try using us in Elastic Cloud, and you'll get the same experience." But we've evolved a lot, and there are a lot more features. And in the cloud, there's no such thing as free because you're paying for the service no matter what. And developers and DevOps users are used to that. So we see that as a pretty strong strength, and it's over 10 years of very rigorous IP development. Like we've always been known as an innovation-centric company. That's been our genesis. And the competitive moat is pretty deep and pretty wide.

I want to ask you 2 more big-picture-type questions, and I do want to open up to the floor to see if any investors have questions, too, for you. So the 2 questions. One, kind of on the macro, right? I want to kind of go back to the big picture here, the demand environment overall. It definitely sounded like on the last earnings call, demand is great out there, but there's a lot of things going on in the world. There's disruption in Europe. There's this war going on, inflation, recession risk, multiple places. I mean, have you seen any impact to the demand environment at all? And are you -- how are you guys thinking about positioning the business, if at all, for any sort of potential recession out there in the future?

So in terms of the demand environment, we have not seen any change in the demand for our offerings. And the reason why I've been on the road is, to me, it always helps me to get a first-hand view. We obviously track a lot of signals from our field. And in all the data points that we have, like I said, like the stickiness of the solutions gives us a lot of confidence in the demand and how people will prioritize their spend as they need to prioritize it. We are not blind to the macroeconomic environment. So obviously, the inflation risk everywhere, like all of that is something that we track as well. Now in terms of the other impacts, obviously, the labor market is something that we track very, very carefully. Us being a distributed company gives us a lot of advantages. I don't know if most people are aware of this, but Elastic is -- we are completely distributed. We don't have hubs in a way. So we have employees in over 42 countries, the last I checked. And what that lets us do is our philosophy has always been hire talent wherever it lives. And that, in itself, ends up being a wonderful hedge against the kinds of pressures that you see in certain talent markets around the world that sometimes get too heated, sometimes you see other impacts. So we are able to manage our overall costs and our overall attrition very, very nicely because of that. We're able to hire where we need to appropriately as well. We have always been very thoughtful and disciplined about our expenses and where we invest. That's always been something that we've been very conscious of, and we will continue to do that. So it's a -- I think in the last earnings call, we also talked about the fact that, over this multiyear period, in FY '24 and FY '25, we will be growing our margins by several points. Janesh always reminds me that when he started at Elastic, our operating margin was, I think, negative 20%. And this past year, we ended at effectively slightly positive, basically flat operating margin. So we've seen the ability to both grow our top line and improve our margins over time, and you should expect that we will continue to run the business with that same kind of discipline. But we believe in building a business that has durable growth but also profitable growth. We've always done that, and we will continue to do that.

That's great. I want to follow up on a comment you had, and I was going to save the question for later. But you talked about it, so I'll ask it right now. It's about the hiring environment. You guys are distributed. When we think about software -- the category as a whole and actually the entire world, I guess, 6 months ago, everyone says they couldn't find enough people. It was a very tough hiring environment, but things have changed. So how are you guys viewing the hiring environment for Elastic, specifically today and maybe over the next 6 to 12 months?

We have not seen a change yet. Now given the fact that hiring might be slowing down in certain places, we are paying careful attention to what that might mean and if the markets -- if the hiring environment becomes easier. But at this point, we have not seen any change in the hiring environment overall. It might also be because of the fact that we are as distributed as we are, and we've never had those kinds of pressures that come if you are entirely located in the Bay Area or entirely located in New York or London, like some of these talent hubs tend to be. They tend to have wilder swings, if you will. Like we've been a lot more distributed than that.

Okay. Last big-picture question before I open it up to the floor. Competition, competitive environment. Elasticsearch, very, very best in class almost, right? How do we think about the competitive landscape? Or how do you guys think about the competitive landscape between Elasticsearch that you're seeing products, and then, I guess, the most topical is Observability. So how do you categorize it for yourself?

Yes. So first and foremost, it's important to understand that we've always approached this as us being a data analytics platform that's powered by search. So we are first and foremost that, and we are used for Security, Observability and Enterprise Search. And over time, in the future, there might be additional solution areas that we enter into. Because we have that flexible platform, the reality is that we didn't enter into security. We had customers lead us to security. So we saw customers using us for security use cases for SIEM, and that's how we got into the market. And we always track what else are we being used for, and those are potential future opportunities for us. Today, we are in these 3 in security. In SIEM, that's where we are the most mature. We have had our SIEM functionality for the longest period of time, and that's where we have powering strengths. We are very confident we lead with SIEM. And the way we approach the market is with this notion of let's help bring visibility to your environment. Let's help you with the detections, first, to detect the threats where they are. And once we get in that way, that's the land motion, then the expand actually becomes, hey, now that we've shown you where the issues might be, can we help you protect those environments? That same Elastic agent that is pulling data from that agent, from that endpoint, we can now use that same agent to start blocking ransomware, to start preventing malware attacks, to start showing you if there is somebody trying to do any kind of privilege escalation. And that's a huge strength for us because the architecture is completely vertically integrated, so it's not like you're using one solution and you need to now install a different solution, and there's a new binary that you need to install. You just -- once you have Elastic install, we're just turning on features. You just start using more and more features, and it's as simple as that. And that is a big strength we expect to continue driving that momentum. We announced cloud security yesterday at RSA or a day before yesterday, whenever that was. And that becomes the next area now because cloud workloads are primarily Linux-based workloads. So we announced cloud workload protection and Kubernetes Security Posture Management. That's going to be the next expansion after XDR that we will continue on, on Observability, which we talked about. We lead with log analytics. And so on log analytics, again, that's our core strength. You can imagine who the primary competitors are in that area. And over time, we've been expanding from there. So we land with log analytics, and then the value proposition becomes, hey, you're using APM. You need APM. Why don't you start using our APM agents rather than using whatever else you're using and start instrumenting a few applications to begin with? And over time, you'll see the value of having all of this data in one data store, where you can start to do correlations. You can use machine learning jobs, to using ML, what people talk about, AIOps, identifying issues early, proactively bumping them up, so the SRE can do something about them. Like all of that just becomes easier if you have this vertically integrated stack. And with Enterprise Search, we always start with App Search, that ubiquitous search bar that you see. And from there, then we say, "Okay, you've already done that with one data source. How about expanding it and getting into Workplace Search where you're pulling in data from multiple different data sources?" So it's a land and expand motion. And although we have competitors in each of them, we are focusing on where we have the greatest strengths and expanding from there. And we see that as an approach that gives us a lot of advantages.

Thank you. Any questions from the audience? Please raise your hand, and we can get the mic over to you.

Exciting announcement a couple of days ago on, I think, cloud security. So a lot of vendors doing that right now. So how do you position yourselves versus them? And some of them are, I guess, partners. So what's the pitch there versus Wiz, Orca, Palo? That would be great.

Yes. So for us, I think the way we've always led into security is, first and foremost, like I said, we'll always lead with visibility, right? So we always lead with SIEM. And in terms of differentiation on the functionality, first and foremost, it's fully integrated. So you don't have to use another agent. It's the same agent that's being used for giving you visibility into your Kubernetes environment for observability that now allows you to secure those Kubernetes workloads. We acquired a company called Cmd last year that had a lot of experience and expertise in Linux security. The entire endpoint industry -- endpoint security industry has grown on the backs of Microsoft because the most -- the least secure operating system environment was Windows. And so most of the traditional vendors have all had experience in Windows security management. Linux is the primary OS on cloud. And so when we started to get into this area, we acquired a company that was a specialist at Linux security. Because those threat vectors tend to be different, they tend to have different challenges, you need to approach it differently. And we acquired that company. It's -- a lot of the capabilities that are built on this technology called eBPF, and eBPF allows you to really get amazing visibility into what's happening in the Linux kernel. And now all of that has been integrated into our platform. So you can, from one interface, see all of your data because that's your SIEM and from your SIEM start to now get visibility into your cloud security and your cloud posture. If you're using anybody else, you're always going to have your SIEM, and you're going to have a different console here. And the value proposition that we can deliver is if it's the same people who are doing all the detections and investigations and the securing, why make them work on 2, 3 different consoles? We can show you how this is a much natural fit. And since they already have SIEM, Elastic SIEM, it's an easier move for them. So we are -- I'm very excited about the opportunity here.

I had something of a similar question. If you look at some of your competitors like CrowdStrike and SentinelOne, they've talked about their own threat hunting and same XDR product. CrowdStrike also has Humio, which is actually logging. And then recently, SentinelOne actually talked about a data module which does more of the data analytics and stuff and specifically mentioned your name. So it seems like in the last few months, the space has been heating up in all these different directions. So if you can address that.

Yes, absolutely. So a lot of the vendors that you talked about, they've all come at it from what people refer to as XDR, right? An XDR has really grown from endpoint technology. It's -- everybody -- all of those vendors started as endpoint security vendors primarily focused on windows. And from there, they've gone to add more and more modules and so on. And their -- XDR is effectively I'm going to start with my endpoint data, and then I'm going to try and expand from there. And what that naturally limits you to is how flexible is that data model and how much data can it allow you to bring in and truly analyze. And there is a fastly gaining acceptance in the market, in the security market of the fact that XDR is never going to replace it because a SIEM is by design an extremely flexible data model because you want to bring in network data. You want to bring in web data. You want to bring in your data protection information. You want to bring in your auth information, your endpoint information. You even want to in bring in your physical security sensor information. And XDRs just aren't -- they didn't start with the notion of an analytics platform. That's not how they started. They started with endpoint security. Now they're bolting on things to make it work, and we see that over and over again. I have -- it's fun to see that XDR is the same as SIEM. You're always going to have SIEM. You might move to an XDR. We see it as an extension more than anything else. And they are approaching it from the XDR standpoint. We start with SIEM. We establish ourselves, and we say, "Okay, now we can also help you protect, so we can do the -- not just the detection, but the R in, whatever, DR." And that's the approach that we take. And because of the predominance of SIEM, when you look at a SOC, a security operations center, really what they worked most with is a SIEM. That's the cornerstone because that's where all the data eventually ends up. It's a very nice place for us to grow from there. And when it comes to scale, Koji said it. I didn't say it, but I agree with him that there is nothing out there that we have seen that deals with scale like Elasticsearch does and performs the way Elasticsearch does. So we feel very confident about our position.

Ash, I wanted to ask you a question on the license change that you guys implemented, I think about 18 months ago now.

Yes.

It's about 1.5 years later, right? So it's been 1.5 years since you've changed it to an SSPL. Any thoughts on kind of feedback from the community 18 months later? How do we think about it? Any sort of update on it? I mean, anything on the SSPL change and the license change?

Yes. It went exactly as we expected because the way we crafted the license -- and it wasn't just SSPL, but we also created Elastic License v2, and we put all of it in both Elastic License v2 as well as SSPL. But it was designed to be a very permissive license. We wanted people to have the ability to do what they used to do in the past as well, which was take Elasticsearch under the covers, use it if they wanted to build their own application. Basically, what it said was you can't take Elasticsearch, run it as a service and then monetize just that as a service without building your own IP on top and building your own application logic, which basically is different from Elasticsearch. So it was, for a lack of -- and just to be clear, it was targeted towards stopping one form of bad behavior that we were not very happy within the market, right? So -- and we accomplished that goal. Absolutely. We still have lots and lots of users out there, companies that use Elasticsearch under the covers, continue to do so. Many of them use Elasticsearch for free, legally under Elastic License v2. We are super excited about that because, someday, they will become prospects for us as we deliver searchable snapshots. This capability that we delivered about 18 months ago that makes it possible for you to use low-cost object storage. And when you use that low-cost object storage, your storage costs decrease significantly. You're able to store more data. And as data volumes are growing, we expect that those kinds of customers that might be using the free version today will be motivated to talk to us about the paid version in the future. And if they use Elastic Cloud, there's no such thing as free. So I think it helped us make sure that we brought the community along the way we wanted to while stopping what we wanted to stop. We've always been very confident about the product differentiation that we've had versus what AWS had as their service. They are now -- as of now of 7.10, which is over a year old. The pace of innovation that we are able to drive is at least an order of magnitude greater than them, if not more. And you can see that from the pull requests and other things that allow you to see how quickly things are evolving. And so we feel very, very good about where this has placed us. And honestly, it has improved the relationship between the 2 companies, between AWS and Elastic, because now there is clarity in the market. Now when an AWS seller gets inbound for Elasticsearch, there's only one Elasticsearch, and that's from us. And they always got some quota retirement for selling things that were in the marketplace, and now there is no confusion. There's one Elasticsearch. That's from Elastic. There might be this open search thing from AWS, but those are 2 completely different products. They don't have any observability or security capabilities. That's been the fact for the last many years. So it's cleared the air. It has allowed us to partner better with AWS. Their -- I can't remember whenever their executives have sort of talked about their collaboration with AWS -- with Elastic. So you see all of those signs of how the relationship has improved because of some of those changes that we made. So I'm personally very happy about it. There's motivation on both sides, in AWS and us, to work together because there's a massive install base out there and opportunity out there. So in the long run, I expect that to be a really good thing for us.

Ash, thank you so much. We are all out of time.

Wonderful.

Thank you so much for this presentation. It was great.

Thank you very much. Appreciate it.

Yes.