Okay, cool. Why don't we get started, guys? Welcome to day two, afternoon of day two, of the flagship TMT conference. My name is Erik Woodring. I lead Morgan Stanley's hardware coverage here. I am delighted to be joined by John Ederer, CFO of Teradata, Sumeet Arora, Teradata's Chief Product Officer. Both you guys, thank you. Absolutely. Welcome to the conference. Thanks for having us. Before we, before we begin, let me point everyone to the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. With that out of the way, you know, this is the first time that I've been able to host both of you at the TMT conference, both relatively new to the role that you're sitting in right now. Yeah. I'd love just to start the conversation, just quick background, maybe one or two priorities for each of you as you think about the changes you want to kind of enact in your relative seats, and then we'll go from there. Sure. Well, I'll kick things off, even though Sumeet predated me by about 30 days. Yeah, I joined last May, and very excited to be on board. I've been in the software industry for 20+ years. And before that, I actually started in your chair. I was a research analyst for about a decade before making the jump over. Spent a lot of time in enterprise software, most recently with a company called Model N that was publicly traded and then joined up with Teradata last spring. You know, as I came on board, we, as a company, had obviously gone through, I would call it a challenging 2024. A big priority for us in 2025 was really just kinda getting the business back on track. There were some things that we felt were really critical. One was getting total ARR back to positive territory. The second was continuing to improve on our operating margin. Third was really starting to demonstrate durable free cash flow generation. I think fourth, and perhaps maybe the most important, was just demonstrating some consistency on a quarterly basis as we went through the year. I think we were able to do all of that, and that was probably goal number one coming out of the gates. Cool. Sumeet? Yeah. I'm Sumeet Arora, I'm the Chief Product Officer at Teradata, joined roughly around the same time as John. I have 30+ years in tech industry, have spanned domains. I did Cisco for 20+ years, built some of the largest routing systems that operate the internet today. I have spent more than five years at a much, much smaller company focused on agentic analytics, natural language analytics, way before ChatGPT became popular. In the last 10 months, I've been at Teradata. My priorities, one or two priorities. I think first is to make sure we continue to earn the trust of our customers through our product. That's most foremost. Second is really focusing on innovation, both foundationally as well as, in some ways, almost like a startup. You know, operate like a startup in terms of the velocity, in terms of discovering product market fit for some of the newer things that are happening. It's kinda almost two-faceted there, and those are my priorities. Okay, great. No, that's a great way to start the conversation. Sumeet, not everybody gets exposure to you, so I wanna take advantage of this and start with you. Maybe, again, starting at a high level, just the demand landscape as we look at Teradata solutions, especially how they could be changing with generative AI, risk or opportunities around agentic AI, you know, what are the solutions that you see right now, you know, kinda specifically gaining traction with customers at Teradata? Yeah. I think, the landscape, I'll just say that the landscape above data platforms like Teradata is shifting. From traditional applications to reporting, which are still there, but there is absolutely a motion towards agents, agentic AI, experimentation with that, actual stuff. I can see that just based on the demand for MCP servers and certain types of queries that show up, right? I can see that shifting right now. That's kinda part one of the story. The part two of the story is the demand for context. Everybody's talking about that, but what is context? Context is the ability to give a worldview to agents so that they can be more enterprise-grade, explainable, high accuracy. Teradata, platforms like us, have the knowledge of the data, the metadata, also the knowledge of the industries that we have served for more than four decades, and our ability to combine these to deliver the right context for the use case for the agent, so that agents are useful, actually useful, deliver real ROI, is the second piece that I'm seeing in my conversations. The third piece is, people want AI to be co-located with the gravity of data. There is no AI without knowledge, without enterprise knowledge, enterprise data. We are known to be that trusted repository. People want to locate AI where that data is instead of moving the data to where AI is. That is an amazing opportunity. I see that in all my conversations. I'll just say that underlying all of these three that I outlined is the demand for moving fast with innovation, but minimizing risk. The moving fast with innovation, but, like, having cost efficiency and cost economics. Moving fast with innovation, but really having the governance and the trust. I think Teradata is really well positioned. Okay. To deliver on those. Yeah. Perfect. If I could just tack onto that quickly. You know, if you look back at our business over the last, you know, three to four, even five years, you know, we were very focused on building the cloud side of our business, and I think we demonstrated a very strong ability to do that. The emphasis has shifted and whereas even just a few years ago, it was still very much about migrations and getting to the cloud just for the sake of getting to the cloud. That emphasis has really shifted to some of the things that Sumeet was just describing from an AI standpoint, and there's an eagerness to get going on those investments, and using the data where it sits today, whether that's in the cloud or on-premise. Okay. No, good. We'll get into that kind of Teradata 3.0, so to speak. Yeah. I wanna touch on competitive landscape. Before we get there, I'd be remiss if I didn't kind of let you share your thoughts on AI disruption, right? Obviously, kind of a key topic here, key topic in the market. You know, Teradata should play an important role in enabling AI. There's broad risk or concern that new LLM tools can kinda disrupt existing software platforms. Just maybe lay out how you think about the risk of AI disruption or the opportunity with AI disruption just as it relates to Teradata. Absolutely. From a Teradata vantage point, if you look at the new enterprise tech stack that is emerging, there is essentially a knowledge fabric, which is the data, the processes, the documents, the important videos, and maybe what's in the heads of people. That's the knowledge fabric. That's the truest set of an enterprise. There is a layer that essentially uses that to deliver governed context, trusted context. There is the agents layer, which is the layer that is kinda approaching or trying to replace the SaaS layer, right? There is the interface with the stack itself, which is going outcome-centric. If you look at that stack, you look at Teradata, we are absolutely well-positioned to play the role of the knowledge fabric- Mm-hmm ... and being part of that, both with our structured data capabilities and our absolutely first, you know, first-class, vector capabilities, because we can understand unstructured really well. we are a big part of the knowledge fabric, and I think as long as I and my team and Teradata, we make sure that this is usable by agents, which is this layer above, you know, trusted context, which we are absolutely invested in at Teradata, delivering that to agents. you know, people can build agents in Teradata now. They can deploy agents close to literally in the Teradata environment. They can govern them, or if needed, they can build agents elsewhere, but they can come to us through the MCP Server. we offer both-. Mm-hmm ... because we want that agentic layer to have the flexibility to run on-prem, like John was saying, run in the cloud, and safely access on-prem data. We are really enabling the agentic layer as well, both in terms of supporting other parties, but also first-party agents inside of Teradata. We see us playing a huge role in at least three layers of this stack, all the way delivering on trust and governance as the primary property. Okay. Let's touch on kinda the competitive landscape and Teradata's differentiation. 2024 was a year of higher than average erosion. You've been able to bring that under control. John, you talked about kind of executing in 2025. Yep. I know your focus is on hybrid workloads also continue to move to the cloud. I'm just kinda setting the stage for if we think about these large scale workloads where Teradata excels in price for in performance, is that the impenetrable competitive moat that you believe Teradata has or, you know, across kind of product pricing, performance, et cetera, where is that impenetrable moat? Yeah. Maybe I'll take a shot at it. Yeah. Okay. I think the first piece is our ability to give choice to our customers. They have the choice to leverage the same technology stack on-prem or in the cloud, which is increasingly important. The deployment type dictates economics and security. Mm-hmm. The second piece is the choice we provide in terms of storage options. The reason that's important is Teradata can do both block store and object store really well. The advantage of the block store tightly coupled architecture is low latency and high concurrency, which is very amenable to operational workloads, increasingly being delivered through agents. Mm-hmm. The same engine is built for the always-on economy. Remember, what my customers tell me is Teradata is the best when it the workload is always on or when the workloads are mixed, because we have this amazing workload management technique. Guess what? The cloud story was built with human usage patterns in mind. We go to bed at night, we take time off at weekends, things shrink in the cloud. That delivers economics. Now we are entering a time when agents don't sleep. They're vigorous. They're 25 times more vigorous in terms of the amount of data they access. In that era, that always-on economy, we are the price performance option. That's kind of one choice we offer. We also offer the elastic compute option, which is amenable for things that come and go, things that are experimental. They get choice of two types of compute. Again, price performance. Types of storage, types of deployment environments. I call it as the best engine, the best context. Remember we spoke about context earlier in my conversation, our understanding of industries. Like, we understand financial services, we understand telcos, we understand airlines. Mm-hmm. We can bring that knowledge and map the crazy taxonomy of data that exists underlying to give a worldview to agents that agents can be much more enterprise-grade. Best context, best engine. We built, I think we have already spoken about this, so I can say it, which is we're building this AI studio capability, which is right next to our data, our knowledge, and that allows people to really bring their AI outcomes right there next to the data without data having to move, and that's a beautiful option. These are some of our differentiations that actually resonate with my customers. Okay. John, you were just gonna say something? No, no. Okay. Okay, great. Sumeet, maybe just add on to that. When we talk about the innovation coming out of Teradata, these agentic architectures and protocols that are very rapidly emerging, you know, what's the criticality of the role they play as we think about Teradata innovation over not just, like, 2026, but as you're looking out and you're thinking about how Teradata needs to remain the most relevant for their customers- Yeah. -or their trusted choice. Talk to us about that a little. That's a great question. Look, one of the things that innovators forget is the value of distribution, right? It's important to build great technology. Mm-hmm. If it's not easily consumable and it's not distributed well, then you don't get derive the value, and you're not able to invest in it. If I look at the innovation we are unpacking this year, whether it's complete transformation of our hardware, you know, like, a huge investment in AI, a huge enablement of agentic workloads, it is also what we are trying to do is trying to make it easy to consume, easy to use, and really focusing on some building blocks that allow me to play the market to continuously discover product market fit. I think as companies in this era of AI, we have to almost behave like startups every day like day zero, because things just change so rapidly, Erik. Like, you know, Anthropic drops something, and, like, whole world is, like, reacting to it, right? In this era, you wanna have the building blocks in place so you can play the market based on what the emergent needs are. I think I would characterize my our set of innovations this year as also the building block- Mm-hmm. -for that. I think you're gonna see us do a lot of partnerships in big ways this year as well, which is another way to drive distribution and leverage. Okay. Awesome. going back to the kinda Teradata 1.0, 2.0, 3.0, I've been privy to see all evolutions of that. you know, IntelliFlex boxes on-prem to we're going cloud first, now it's kind of the hybrid approach. Yeah. This could be for you, John, or Sumeet, but just talk to us about why, you know, this is the approach that gives you kind of that long-lasting competitive differentiation, sets you up for the most long-term success? Yeah. All right. If you wanna talk about it from a product standpoint? Absolutely. you know, look, one of the things I get, which is the most pleasureful part of my job, is the opportunity to meet customers and prospects around the globe. Mm-hmm. There is increasing value attached to the ability to deliver innovation wherever it's needed. It could be an on-prem appliance. It could be a private data center. It could be a sovereign cloud, a local cloud. It could be the public cloud. The Teradata tech stack ships in all those locations. Not only does it ship to all those locations, we ship the same stack, the same set of innovations, the same AI that you can do in public cloud, you can do it in all those areas, on-prem, appliance, wherever it is, you can do it anywhere with Teradata, it's not like one lags the other by years. No. Within few weeks, everything is available everywhere. That is the velocity at which we are moving. We're making sure that the best is available in all these locations. Customers love that. Not only do we ship these and deploy these in different locations, they can talk to each other. You have real, you can mitigate risk, you can put the right economics, you can deliver based on privacy and security requirements, and I think we're really, really well-positioned. That is the conversation that happens, around the world. Mm-hmm. I feel very, very strongly about that. Maybe, John, this can be for you. Yeah. As a jump-off point from that is, so what does that mean when we think about the underlying drivers of growth when it comes to, you know, new logos versus limiting erosion versus migrations versus expansions? Yeah. When we talk to this more sustainable ARR growth. Mm. Where is that coming from if we kind of dissect it across those different avenues? Yeah. I mean, you kinda touched on every element there. Okay. I would say what's different in terms of the go-forward view is probably the relative weighting of those things, right? Mm-hmm. I alluded to this a few minutes ago. If you were to go back a year or two, there was a heavy emphasis on migration activity. Mm-hmm. Right? That was used as a, as, one, a point to get customers to the cloud, but two, a catalyst to engage with the customer again and upsell and cross-sell and trying to expand the relationship overall. What we're seeing now is really a move towards the adoption of AI, and that is a primary driver for us. That drives the increased workloads, and that's what fuels a lot of the expansion activity for Teradata. I'd say going forward, what you would see from us is a much more traditional land and expand type of a model, where, yes, we will continue to focus on new logos. We'll also look to expand existing customer relationships. Okay. Perfect. I am just writing that down. Awesome. Now, maybe Sumeet, last question, before I turn it over to John and kind of turn this into numbers is, just going back to kind of the product set, we've talked about MCP Server, you've talked about the kind of NVIDIA partnership that is relatively new. What are, what are you most excited about from a new introduction standpoint? What does this mean for the pace of innovation that's gonna come out of Teradata over the next one to three years? You talked about, I think it was 150 AI and agentic engagement in 4Q. Bring that all together for like... Where is Sumeet getting excited about where I need to push that product to, so to speak? Yeah, yeah. I think, look, this is the year where we are bringing to life through execution, raw execution, our vision of an autonomous AI and knowledge platform. We are moving the addressable data for Teradata from structured data to also making sure that we are able to handle all sorts of unstructured information inside the enterprise, whether it's videos, images, documents, whatever it is that you need for AI. We are evolving to the trusted knowledge platform. That's kind of part one. The second piece is we're delivering AI outcomes. Everything starts with ROI, with measurable outcomes, work backwards from there. We're delivering AI outcomes right where the knowledge gravity is, where data gravity is, and in different environments like we discussed. To support that evolution to the autonomous enterprise, we're delivering our own stack to be fully agentic and autonomous. That's kind of the vision and strategy. What am I most excited about? I think this is a big year for us. I think, if I look out the next few months, up until summer, and so on, we have a full transformation of our on-prem platform underway. We'll transform on-prem to an autonomous AI and knowledge platform. We'll enable people's AI factories with knowledge and help them deliver outcomes, right? That's a big deal. In terms of cloud, it's going to be an exciting year for us, where we combine the low latency, high concurrency, always-on system with ephemeral, elastic compute that delivers both the ability to innovate fast, but also deliver price performance for stuff that needs to be always on and proper and structured. I'm very excited about that prospect, both in the cloud and on-prem. That's the story. And I'm super excited about our AI Studio be launched. Mm-hmm. Which is going to bring together... I know we didn't dwell on it, but I'll take 30 seconds and finish this off, which is, the AI Studio brings the ability to do entire agent lifecycle, build, deploy, govern agents, with the MCP Server. Allows the ability to do all sorts of analytics with 150+ in-database analytics functions that we have. Allows the ability to run applications that are oriented at specific use cases and outcomes. Last but not least, allows for scalable Python compute to leverage both, normal CPU compute as well as accelerated GPU computes in both on-prem and cloud environments. My customers are loving this vision and strategy and the execution, and I hope to bring that to life this year. Cool. I'm, I'll maybe transition more to John now and find out how we're turning that into numbers, so to speak, on financial statements. Yeah. John, as someone that's covered for Teradata for a decade, it's exciting to see stabilization in top line. Yeah. Yeah. A modest acceleration in the business, 2%-4% ARR growth. Yeah. Flat to 2% recurring revenue growth. We've kind of talked through how you expect to get there, so to speak. Yeah. - with expansions kind of primarily, the driver. Yeah. Talk to us about how you see kind of upside and downside risks to that? That can be macro-related, that can be company and product-related, just up and downside risks. Yeah. I mean, I think that. Well, look, I prior to this week, I would say, you know, more focused inwardly and on our own execution. Obviously, there's a lot going on in the world today and that's a very large wildcard for all of us. I think really when we look at how we're operating, how we operated through 2025 and then how we set ourselves up for 2026, it really is about execution for us. We did get back to positive territory on total ARR growth last year. That was an important milestone, especially relative to 2024, to get us back on track and to start to stabilize the business. We did launch some products last year. We've got a lot more slated for this year. We did start to do some things from a go-to-market standpoint to hit the ground running in 2026. We talked about on our earnings call, the 150 proof of concepts that we did on the AI side. We have our forward deployed engineers out there working very closely with customers, developing use cases, creating these POCs. We've added a services team, our AI services team, to come in behind that and take those proof of concepts and spin them up into actual projects and get customers launched and going on this initiative. Those are some of the ways that we're starting to try combine the product side with the go-to-market effort to really start that expansion activity in 2026. I think if I look at sort of the upside and the, and the downside risk, I do think a lot of it comes down to our own execution in 2026. Yeah. Yeah. Okay, perfect. You know, we talked about expansion, again, key driver of growth. You know, if I were to maybe just push on that a little bit, cloud net expansion, that's a rate that has been falling. I realize that cloud isn't necessarily the maybe the KPI that it was during Teradata 2.0. Yeah. This is the hybrid story. Yeah. To be fair, cloud is still kind of the primary driver of growth. Mm-hmm. So to speak. Mm-hmm. When does that rate kind of stabilize? Just walk us through where you expect to see these expansions? Yeah ... happen. Yeah. No, that rate has been, has been shifting as we've moved through, the migration period, and especially if you look at the total ARR growth rate, that was obviously impacted for a few years by migration activity, and now we're starting to come down on the other side of the bell curve. Mm-hmm of that activity. I would say we're close to seeing that normalize. What I would expect going forward is a little bit more convergence between that net expansion rate and the total ARR cloud growth rate. Okay. Okay, helpful. I know you don't guide beyond 2026. I don't expect you to give that. It would be helpful for us to better understand if you see 2026 as kind of an inflection year that should enable a more consistent return to growth or even how, you know, you guys are thinking about the revenue growth trajectory from here. Presumably, the fact that ARR is growing this year would set you up for more growth in 2027. Yeah. Like, where are you on this journey from stabilization to potential inflection? Yeah. I mean, at the risk of oversimplifying it, you know, we're looking at this journey as a bit of a crawl, walk, run scenario where, you know, 2025 was the crawl. That was kind of getting us back to a stable business, getting us back to the positive side of growth from an ARR standpoint. 2026, we're certainly looking to build on that. Even just from a total ARR standpoint, we were at 1% constant currency growth last year, 2%-4% guidance this year. Making some incremental improvement and continuing to build the foundation. The other thing, what we're doing in 2026, and we talked about this on our call, is that we're really prioritizing our investments. We're still looking to optimize parts of the business, so we can drop incremental benefit to the bottom line. Mm. We're also reinvesting some of that. In particular, we're investing on the product side. A lot of the things that Sumeet described and we're working on now, and we expect to launch over the. Mm ... over the course of this year, will become drivers for 2027 and beyond. Okay. And then- Oh, sorry. Yeah, yeah. We're certainly looking to start to drive that glide path. Yes. I'll stop short of giving you guidance for 2027. Obviously, we're looking to continuously improve. Fair. Okay. I wanna touch on. I'll ask you kinda two questions about margins gross and op. You know, just on the gross side, cloud gross margin is still modestly dilutive to the platform, I believe. Mm-hmm. How do we think about, again, these expansions and kind of the hybrid approach? When does that stabilize gross margins? Yeah. Second to that, you know, these initiatives on the cost side, which is reinvesting in innovation, but also finding areas of efficiencies and rationalization. Yeah ... what does that mean for the trajectory of OpEx? Then we can all kind of back into how we think about op margins can kinda go from here. Yeah. We'll kinda walk down the P&L there. You know, first, from a gross margin standpoint, and I think you kinda mentioned cloud gross margins, but I think it's really the total recurring gross margin, which includes both our on-premise subscriptions and our cloud business. Today, the on-premise subscriptions have a higher gross margin than the cloud side. Cloud has been growing faster. Mm-hmm That presents a bit of a headwind to the gross margin. We have been incrementally improving cloud gross margin each year. We don't disclose that externally. We've been making progress. Okay. We're very focused on that side of it. Particularly for 2026 and beyond, I feel like we really need to start making some stairstep improvements there to ultimately drive future operating margin improvement. If you look back over the last several years, we've made great improvement in operating margin, over 500 basis points over the last three years. Yeah. I would say that was largely on the backs of streamlining operating lines, right? We still think there's a path to do that even with reinvesting in R&D in 2026, our guidance would suggest another point of improvement on the operating margin in 2026. I think to really sustain that and maybe even take that, you know, stairsteps higher, we really need to focus on the gross margin side of the equation. Okay. Okay, perfect. We wanna kinda think about, you know. We'll say it this way. Execution is key this year. Mm-hmm. Remains key, is we wanna hold you guys accountable to say, you know, you're executing the way that you intend to execute. Mm-hmm. In addition to what I imagine is ARR growth, that's kind of one of the most important KPIs. Yeah. Is it earnings growth? Is it operating margins? Is it free cash flow growth? What are the other KPIs we all should be thinking about that's like, "Hey, we're gonna hold John and the team accountable."? Yeah. These are the metrics we should be tracking. Yeah. No, I think that's fair. I think that, yeah, certainly, you know, from a top-line perspective, the total ARR number is the more forward-looking number, right? If you're thinking about what's gonna drive growth in 2027, you should be focused on the total ARR growth in 2026. Certainly, I think it's valid and important to continue to see how we're doing on the operating margin. Mm-hmm. Are we continuing to drop incremental benefit to the bottom line? When we talk about profitable growth, that's exactly what we mean. We're looking for ways to invest that will drive future growth, but we also wanna do so in a responsible way that we're dropping that incremental profit down. From there, I would probably focus on free cash flow. Okay. I think that earnings per share is also a good gauge, although, you know, there's some different dynamics at play there. I think if you look at the free cash flow number, I view that as a critically important number. Even again, if I go back to last year, you know, we had given guidance that was roughly $260 million-$280 million in terms of free cash flow for the year, and we had done $277 million in the year prior. Mm-hmm. As I came on board I was, I was very focused on the amount that we had done last year, and was very focused on making sure we got to that level or higher in, in 2025 and was pleased that we were able to come out a little bit better than that. I view that as a, as a critical foundational element for us from a, from a stock standpoint, frankly. Yeah ... from a valuation standpoint. I feel like that gives us a good floor for investors. Just on free cash flow, kind of the underlying drivers there. Obviously we hear about profitability or the expectation for profitability improvements, that's clearly one. Mm-hmm. You've had some restructuring costs. Yeah ... in the model the last couple years. Yeah. Are those rolling off? Do we expect more of them? Like, what are the building blocks to get that? You know, this year I think it's 10% free cash flow growth, roughly is the target. What are the building blocks that get to that more kind of sustainable free cash flow growth as you're kind of outlining it? I mean, certainly the face of the P&L is a big part of it, right? Continuing to drive that operating profit improvement, and looking at even below that, looking at the EBITDA performance, you know, that's your front end driver to it continuing to be focused on the working capital side. Doing a good job on managing collections and payables and those elements. Being diligent with our capital spend, our CapEx. We've generally done a good job of that. There may be some investments we make this year, particularly around the hardware side of the business, you know, in general I think we've done a nice job of managing that. You know, we'll focus on all of the elements that ultimately drive them. Recently, you guys, kind of settled with SAP. You got a $480 million gross payment. Mm. That nets out to, I think it's $355 million-$362 million of- Yeah kind of net cash. Yeah. What do you do with that? That's a good question. Fair question. The honest answer is still to be determined. Okay. I think that, you know, we'll look at our entire capital structure. I think it's strong today. You know, we've got a good amount of cash on the balance sheet. We're generating healthy free cash flow, we're in very good shape there. We do have a term loan about $450 million that comes due in June of 2027, we'll take that into consideration. We're already active in the market from a buyback perspective. Mm-hmm. We announced a refresh of that program, another $500 million, at the end of 2025. We're actively in the market and have committed another 50% of free cash flow going towards stock buybacks. You know, we'll take a look at all of those elements, and we'll talk it through with the board in a few months and. At least leave it at that. Maybe as we think about that, those are maybe... as you've outlined it sounds like those are maybe the most likely. Yeah. I think. obviously. We're gonna look at everything holistically. Okay. Right? You know, you've got, you've got debt, you've got, you know, what are you doing in the marketplace from a, you know, share buyback program. Maybe, you know, maybe there's reasons to preserve some cash if we wanna look at some potential M&A activity. I think all things would be on the table. Okay. With a minute left, I just wanted to give you guys the dance floor. Kind of last comment here. The message that you wanna leave with everyone, maybe what people might underappreciate or undervalue about this story. It can be each of you on it... Yeah, yeah. Go for it ... it doesn't matter, but just... Well, I would just say that Teradata is changing. While keeping the best, of what we have done over the decades, we're changing. We are really enjoying this market in terms of innovation and really operating like a startup in many ways. That's what I would conclude my parts with. Yeah Give you a few seconds. Yeah. I would just tack onto that and say that, you know, it's a business that's undergone a lot of change over the last several years. Certainly the transition from on-premise to the cloud was a big driver of the business for several years. When you're going through those types of transitions, it can get a little muddy. Mm-hmm. The story is sometimes harder to understand. I guess I would encourage people to take a look, and do the deeper dive because I do think a lot of things are starting to tilt our way. I think customers are starting to tilt our way in, and how people are deploying, agentic solutions, again, is starting to tilt back in favor of Teradata. Cool. That's a perfect place to end. Thanks. Thank you, guys. Thank you. Thank you, Erik.