I think we're live. Good morning, everybody. Welcome back. I'm Brad Zelnick with the software team. Welcome to the Deutsche Bank 2024 Tech Conference here in sunny, I think it's still sunny outside, Dana Point, California. For this session, really delighted to be hosting Microsoft, where with us today, we have Corporate Vice President of Business Apps and Platforms, Charles Lamanna. Charles, thanks so much for joining us. Thanks for having me, Brad. Super excited to get a chance to chat with everybody. Awesome. Format of this presentation is gonna be a fireside chat. I've got a bunch of questions, topics I want to run through with Charles, and if we've got time at the end, maybe we'll take a question or two from the room. But with that, Charles, thank you very much. At Microsoft, as a Corporate Vice President, and with a title responsible for business apps and platforms, that could mean a lot of things. So if you can just help level set us, can you take a minute to explain your role a little bit? What is your purview? What's your success criteria? Yeah, absolutely, so kinda the big focus for me really these days is, like, our business offerings and our industry offerings, and when we talk about business, that's really Dynamics 365, so that's our CRM, our ERP, our CDP, so kinda all the classical business SaaS apps. Our low-code offerings with the Power Platform, so, like Power Apps and Power Automate. A bunch of our Copilot extensions and Copilot platform components, like Copilot Studio or Copilot for Sales or Copilot for Service, and then a lot of our industry solutions at Microsoft, and kind of these things go together really nicely because if you look at business apps and industry solutions, they both are similar in that they're very focused on helping our customers transform business outcomes and metrics as opposed to raw infrastructure or productivity. So that's kind of why we have those together. And I do all the product and engineering, so my team, you know, Live Site, servicing, design, kind of that whole thing. Excellent, thank you. That's a great intro to help level set where you sit. Maybe before we dig into all the innovation and business apps, I know investors are interested in what you're seeing in the current environment. In recent quarters, you continued to talk about Dynamics 365 taking share, but at the same time, you know, bookings commentary has been more mixed. Mm-hmm. What broader trends are you seeing? How are customers prioritizing their spend? And are you seeing significant changes maybe in the competitive dynamics, especially with all the focus on AI? Absolutely. So, like last quarter, I think we talked about Dynamics 365 is 90% cloud now, growing 20% constant currency in the cloud, I think about $6.5 billion last year. So mixed gaining share is, I think, a good way to characterize it. And, one of the things that we're seeing right now is that customers are very much looking to make investments around AI that actually produce tangible business outcomes. And I think maybe a year ago, there was more like AI tourism or AI experimentation. Now, the conversation I have with every single customer, every single week, is: "How will this help me reduce my cost for providing service?" Or, "How will this help me go drive revenue uplift for my sales teams or my marketing teams?" And this kind of ROI conversation is great because it... That's how you can go expand the budget customers are willing to spend on technology. So, if you can go make more revenue, there tends to be bigger budgets, which is something we're gonna need through this AI transformation. And the flip side of that, though, is if you can't show that tangible value, customers aren't going to adopt. That means longer sales cycle in some cases, if you don't have clear value and customer references, those types of things. So, so that is, I think, a lot of what's happening right now as it relates to AI, and why you're seeing, I think, a lot of different perspectives from different companies in terms of how much AI is helping them or not helping them, because it's all about how clearly you can make that case for the business outcome and business results. Now, if we look at business applications from a competitive lens, there's not a single customer that I talk to that is evaluating a CRM or an ERP or a low-code platform that isn't starting from how it's going to evolve and change with AI. Like, I don't think there's a concept of a CRM which is free of AI in the future. You inherently are gonna think AI first when it comes to CRM. And if you go look at how changes have happened in the past, say like on-premise to cloud, for a lot of these applications, it was very disruptive in terms of who was the leader and what incumbency provided or didn't provide. We think a similar thing is going to happen when it comes to AI, because customers are going to change the criteria and what they're looking for out of these tools and solutions. So, I think it is a great increased competitive environment. The customers have very high standards in terms of the business outcomes they want, and we feel pretty good from a Microsoft perspective, our role in that going forward. Makes sense, and it's a really important point of the disruptive prospects for AI and what that means at the application layer. Maybe diving into Copilot in Dynamics 365, it was introduced over a year ago. I believe there are now around 10 different solutions, including Copilot and Dynamics for sales, finance, customer service, et cetera, in different stages of availability. Can you summarize the Copilot strategy in a way that we can understand it? And in terms of the different capabilities you're building across the Dynamics portfolio and how you think about pricing for value with AI. Yeah, I mean, the first thing I would talk about, kind of why there are those ten solutions and why customers are using them is, Like, the model itself doesn't really know anything about a company's business process, doesn't understand how do they close their books at the end of the quarter? How do they go do a sales pursuit? What does a marketing campaign look like for them? That's all private data. That's private workflow, private experiences that each company has, that wasn't available to train the model in the first place. So we've been very focused on how do we take these large language models, you know, GPT-4o or what have you, and how do we enrich them with business process understanding, and also enrich them with the private data and private workflows for our customers. To do that, you have to have these, say, Copilot for Sales offerings with customization that the customer can do on top of it to really go help transform the process. Because if you look at the AI, it's not one size fits all. There's not gonna be one thing that you can do out of the box and transform every company at once, and every discipline, and every role, and function in every company. Instead, you're gonna have to take that one product and customize and extend it. So what we've done at Microsoft is, we've spent the last twelve months kind of getting information on this, is we have Copilot for Microsoft as the base. That's the horizontal offering. Everybody that has Office software today or productivity software today should go have Copilot for Microsoft or an equivalent. Then we've layered on top of that these Copilot for Sales, Copilot for Service, Copilot for Finance, or in Dynamics and so on. The benefit being, you're taking that horizontal, out-of-the-box value and then making it understand the specific business process and the specific needs of each and every customer. Kind of that stack of Copilot for Microsoft, these business-specific extensions, and then Copilot Studio, which is how a customer can tweak and configure the thing, is how we can go make it so it actually delivers that business outcome that I was talking about earlier. Because generating like an email faster is good, but generating an email which understands your CRM data, understands your sales process, and speaks to that particular customer, is far more compelling, gonna produce better business outcomes. If we go look at the, like, broader story, that's really what every customer is going to be doing from an evaluation point of view: What produces the best business value for me? And to get there, you're gonna have to have these specific offerings. And we've been really impressed with the usage of this over the last 12 months. We now have over 2 million active users leveraging these Dynamics, basically the Copilots that my team works on, which is different than the Copilot for Microsoft or GitHub Copilot numbers. Over 2 million already using it, regularly, actively engaged, because the value's there. And as we go think about how all of this shakes out over the next couple of years, we think customers, and companies, and app vendors are gonna have to keep going back to: How do I make the AI speak your business? And that bridge is where most of the innovation is happening at the app layer today, is by injecting it into the workflows and apps people use every day, and also inside of the workflows and data processes that they use every single month to run the business. And maybe just, like, one last thought on it is, like, we have some great stats in analysis that we've done to see what is it like to be an office worker. You know, what is it like to be a sales rep? What is it like to be a customer service rep or a finance professional? The typical person is switching applications 17 or 18 times a day. I mean, so there's not one single app or one single experience where I get my job done completely, but the apps that people are using most commonly are the Office apps. So a lot of our focus when we talk about Copilot is: How do we use Copilot to bring more data and more workflow into Microsoft, into Office 365? That's kind of how it shapes out for us. Really helpful. As we think about the utility today- Mm-hmm ... and we think about Copilot, it's really a brand new... Outside of GitHub, Copilot within Microsoft- Mm-hmm -it's a new phenomenon, within the last year, not even a year old. Yeah. When you talk about the difference between crafting an email to really understanding your business, is there any way you can help simplify, contextualize for us? Mm-hmm kind of, you know, the advancement we've made over the last six, nine months. Mm-hmm and where we're going and, you know, what that rate of change looks like? Yeah. So I'll do my best not to nerd out too much about that, but I think it's a very exciting space, 'cause there's a few layers to it. So you have the core model, like the big large language models, like GPT-4o and similar. And then what we're seeing is, we're doing fine-tuning work specific to business process or domains. So think about fine-tuning for a sales process or a I don't know, like classic Know Your Customer banking process, those types of things. And we layer that on top of the model. So that makes the model more aware, it requires less prompting to behave the way you want it to for that process. But that's kind of insufficient to get the things that we want, because if you think about, like, your customer records, those are- those change too often to put in the model itself. You can't put that in the training data, because by the time you train it, everything's changed again. Like your leads, or your opportunities, or the support cases you're working, those are so mutable and so actively used, they're not gonna end up inside the model itself. So that's where we, we talk about this idea of a Copilot system, where we're able to take the models with fine-tuning on top of them, and then marry it together with customer data in real time. Basically, zip these things together so that when you're generating the email, or when you're doing bank reconciliation on the back office, or you're working a support case, you're able to draw from all of these assets in the context of where the user is working, so in the app, whether it's Outlook or somewhere else, to pull all that together to produce the best possible outcome. Because these models are so hungry for context. The more context you give it, the better they perform. The more relevant information you give it, the better they perform. So there's a whole lot of engineering work and magic to take the user context, the data context, the workflow context together with the model to produce the best possible outcome. And if you think about, like, all the product making that's happening in the world right now above the model layer, it's all about basically getting that context right, getting that signal right. Because you can't put an infinite amount of context. And so anyway, so that's kind of the work that we're doing. And when we talk about, like, the business or industry copilots in my team or the copilots inside of Microsoft 365, they're all about getting that context right. And that's where we think at Microsoft, we have a really unique advantage far beyond just the model itself. 'Cause if you think about a typical work environment, who has the most context about what a user's doing each and every day and has access to the applications they use each and every day? It's gonna be Windows, Outlook, Teams, Excel, PowerPoint. So, like, as you go and kind of go look at why we are very bullish, of course, on Copilot, context is gonna be everything, and we have, we think, the best context right now. It's exciting what it can do today, but it's even way more exciting what it's gonna be able to do in- Yeah one, two, however many years from now. Yeah. Thank you very much for that. Maybe if we could pivot to Power Platform, which includes Power BI, Power Apps, Power Automate, Power Pages, Copilot Studio, which I know we were talking about before, is really an important focus. You know, it's growing rapidly, is widely used by customers to drive business process transformation. Can you help frame the TAM for Power Platform? 'Cause I believe you've already said that it generates billions of annual revenue- Mm-hmm and has 48 million monthly active users, which is massive. How do you approach the go-to-market within the breadth of Microsoft's portfolio in a way that not only boosts adoption, but also captures the value that Power Platform creates? Absolutely. Just, like, a little bit of context for folks who maybe aren't super familiar with Power Platform, the big value proposition there is how can we provide a visual or natural language-based experience to help you build dashboards, to visualize data, or to create web or mobile applications, or automate tasks and processes, or create websites really focused on unlocking a broader set of users to become developers. Of course, developers can use it too to build things faster. If we look at the usage and what the addressable market looks like, 48 million monthly active users, we think that the upside long term can be 10 times that. Because if you're an office worker or an information worker, today you send emails, you build PowerPoint presentations, you build Excel spreadsheets. Tomorrow, you should be able to build apps, automate tasks, visualize data, build kind of whatever you need to get your job done, and just like how Excel took calculations and tabular data from something that was done by specialists in a central team and democratized it and pushed it to every end user, Power Platform is aiming to do the same exact thing for a lot of these common developer artifacts. So when I say ten times, that's basically saying the same set of people who use Office at work is the same set of people that can ultimately use Power Platform at work. Now, we're early in that journey. Like, we grew 40% last year to get to that 48 million number, and we think over time, like, that can continue to grow and reach. The growth curve is a little bit different than, say, something like Teams, just as, like, as some context, because it's not just like I send you a link and you use it. Someone has to build something, then people have to use that thing. So there's a little, like, the growth rate is a little bit different. But the good news is, on the other side of that, I think there's actually a lot of monetization opportunity, as we're already seeing with the billions of dollars of revenue being generated. So that's kind of how I think what the TAM looks like and kinda what that is doing in the market. And over time, the reality is, like, every customer I talk to, their demand for creating digital solutions is basically bottomless. Like, they're... I don't think I've ever met a CIO or a CEO or a CRO that said, "You know what? I have all of my digital needs met." You know? They probably don't even get a quarter of the way down their list. One of the easiest ways to go capture all the opportunity by creating these solutions is to empower the entire organization to create these things. And we have some amazing stories out there of companies like Shell, who have saved tens of millions of dollars of real bottom-line savings in the first year of using this tool by empowering what they call DIY or do-it-yourself developers throughout the business. Or, if you kinda go across the board, you see the same thing, even at the smaller level, like Heathrow Airport doing the same thing, you know. So whether it's like, say, small, medium to the largest enterprises, really unlocking a whole bunch of potential by getting people to develop these things. And as we look to the future, we think AI only accelerates this, this whole story, just because it unlocks and simplifies development even further. Makes sense. Charles, there are moments in time. I've been a student of the software industry for a long time, and there are certain statements that get made. In fact, we had Eric Schmidt here earlier this morning, who made reference to the notion that, you know, we sometimes overestimate what's gonna happen in a year, underestimate in a decade, which, by the way, I always attribute it to Bill Gates. He attributes it to Isaac Asimov. I didn't realize that, so I learned something today, if in fact, he's right, and I'll give him the benefit of the doubt that he is. But I heard Satya say something earlier this year about generative AI, and made the point that, you know, for decades, we as humans have had to understand the computer. Now we're entering into an era where the computer understands us. Yep ... which I think is a powerful notion. So I guess as we think about Power Platform, does the infusion of generative AI and natural language into Power Platform help to accelerate the strategy? And can you maybe highlight how you've embedded AI in the platform and any early usage patterns that you're seeing from customers? ... Yeah, absolutely. And the thing that we see with Power Platform is basically every year it's a constant battle to make it more accessible to more people, to therefore grow the number of users and grow the business. And I think generative AI is the greatest gift we've ever had to these types of tools, because it's exactly as you described it. In the past, you had to go read a manual or take classes to learn how to do the right incantations to make the computer do what you want it to do. That is decreasingly going to be the way that we use these tools. You'll use natural language, you'll use images, you'll use metaphor, you'll use example, you'll even speak to the thing, and it will go develop solutions for you. Like earlier this year at our Build conference, we announced this feature for our RPA offering, Robotic Process Automation, called like Record with Copilot, where you literally share your screen with Copilot and you verbalize, you speak to the thought process about what you're doing. So you'd say something like, "Once a week, I go to this file share. I download this Excel spreadsheet. For each row in the spreadsheet, I upload it to this application, and I map these fields to these fields. And if this is true, I click this button." I could explain that to another person. That's how I train them if they'd, like, join my team to do the job. But now I can just tell Copilot that, and then Copilot will build the automation automatically. In that case, it's looking at your screen, so computer vision. It's listening to your voice to provide context on the thought process. It's understanding the context of the applications you use to go build this thing. That is something that starts to be approachable to the typical average user. So we think this goal of Power Platform, where everybody can be a developer, like, really everybody, if you can build a spreadsheet or a PowerPoint presentation, you can be a developer with Power Platform. We think it's more attainable than ever because of these changes, and it's not just gonna be limited to Power Platform. If you look at what GitHub Copilot's doing with Copilot Workspace, you can use natural language and build plans, which will then generate the code and solutions to, say, resolve issues and tasks. There's a real shift happening, where we are gonna have less domain-specific language, less, I think, complexity and sophistication of these tools, and instead, be able to use images, video, audio, natural language to go create these things. That's kind of the big thing that we're seeing right now, and if you go look at a lot of our announcements over the last six or nine months, you start to see that show up. I know earlier, you talked about, like, the distribution and how Power Platform plugs into the bigger Microsoft story. It's been a great example of bottoms-up adoption. These are users who are signing up and trying these things, not being sold from IT down. It's end users going and looking for a solution and using it. And it's done that largely in partnership and extension of Microsoft 365 and our Office applications. With GenAI, we think that boundary that exists between, say, productivity software, development software, business software, that boundary really starts to erode because if I'm using natural language in Microsoft 365 to build my presentations and my spreadsheets, it's not a very far step to also build an application or an automation, or even to start to run my sales process. So kind of, like, just going back to that, we think that there's this big re-shifting, reshaping of where workflow is done, and even the importance of applications. And where you go to start your day will be extremely important because you'll be able to use natural language to do all these things. Now, if we look at kind of how this evolves going forward, the models are getting better all the time. The engineering that we do on top of the models are getting better all the time, and both are important to be successful. And we think that we're earning the trust from customers and training users that you can talk to computers this way more and more, which is gonna be a key part of all of this, too, the change management. Even, like, you're gonna have to re-educate people that, you know, it is okay to talk and search for information with natural language, whereas, like, if you did a Google search in the past, it'd be like, it's really weird to write out a question, right? So you have to do the keywords and stuff. You had to learn how to do it, so you have to unlearn some of these patterns. Very interesting. Copilot Studio seems to be on a real tear. You've talked about 50,000 organizations that have used Copilot Studio to customize or build their own copilots. How does it fit into the overall picture for Microsoft, given the breadth of Copilot offerings across Dynamics 365, but even Microsoft more broadly? Yeah, I mean, the big mantra, like I mentioned earlier, that we use internally is, like, it's not one size fits all. There's not gonna be one model that solves everything for everybody on day one. You're gonna have to have layers of application extensions and workflow extensions, and then have which are specific to departments and industries, and then every customer is going to do a little bit of configuration on top, which is very common. That's how, like, business applications work today. No two customers have the same exact CRM implementation, right? They have their own fields, their own processes, the things that make them unique to their sales domain, so the way they run sales pursuits or the way they do marketing campaigns, and so on. Copilot's gonna be no different. It's not like every, say, CPG company is gonna run their sales and marketing processes the exact same way. They're gonna have some need to customize. And what Copilot Studio does is it makes it possible and easy for every company, in a no-code way, to tailor Copilot to their data sources, to their business processes, and to the content that matters. And inside of Copilot Studio, we have integrations with over fifteen hundred other systems... companies and solutions from Salesforce, SAP, ServiceNow, Workday, AWS, G Suite, if you name it, it probably can be connected to, because we have to bring that context in. And we have a great experience where customers can configure content and prompts that can be used as basically more guidance to these solutions. And, like, one of my favorite examples is we use Copilot at an incredible clip inside of Microsoft. Of course, you would hope that that's true. One of the areas where we've had great success is around customer service, and we have a Copilot for all of our customer service reps. That Copilot's been extended and configured using Copilot Studio to have over 40 different plugins and over 400,000 pieces of content, and it dramatically improved the efficacy of that Copilot. We think that whether it's Copilot for Microsoft 365, Copilot for Sales, Copilot in Dynamics 365, our customers are going to have to configure it using something like Copilot Studio to get the full maximum benefit from it. That's kind of the extension component, and then a lot of people are just building their own Copilots with Copilot Studio. So some great stories, like from PG&E, where they have, like, an HR Copilot called Peggy, which they built in Copilot Studio, and you can go to and ask any questions about HR, and that saved them millions of dollars already in terms of simplifying the process. Or for customer, consumer-facing Copilots, building the same way. So if you want to build a Copilot, Copilot Studio is a great way to do it. You want to extend a Microsoft Copilot, Copilot Studio is a great way to do it, and that's why the growth rate is so sharp and the install base is so large for those reasons. Makes sense. We were surprised. Several weeks ago, we had a conversation with a CIO of a Fortune 500 company, and their vision is ultimately a single Copilot, not just for HR, for- Yeah ... the primary interface for all employees, for all knowledge, for all business process. Yeah. It'll be amazing if we finally get there, too, but why not Microsoft upon which that to be built? Yeah. I mean, it only makes sense. That's where work happens. With respect to time, I got a few questions about Dynamics. Okay. We'll see how much we can get through. Just shifting gears a bit to core workloads in Dynamics 365, there's traditionally been a couple of themes that I'd like to get your thoughts on. Number one, there's been a view that Dynamics 365 is a good option for, you know, smaller businesses, mid-sized enterprises, but not so much for large enterprise. We've actually been hearing of increasing efforts to push further upmarket in some of our industry conversations. Curious to know the latest on how you're approaching this segment of the market, and two, at a product level, the strategy with Dynamics 365 seems to be centered on taking a much more modular approach for focusing on the workflow even. Why is this the right strategy for Microsoft? And does that then open the door to more full-scale replacement of some of the, you know, the larger, well-known names- Mm-hmm ... when you think about the largest enterprise ERP- Yeah -or CRM systems? Yeah, absolutely. So I mean, the first thing is, I would say for Dynamics 365 at this point, we view it from 10 employees to 200,000. 2 million employees, we think, is the full range and addressable market for us. And we have amazing logos that we share every quarter that we put on the board for CRM or for ERP, like, companies like Starbucks and Mercedes and, HP and HSBC. Like, kind of. And these are all, like, relatively recent, the last couple of years logos. And all the time we come out and share just how they've been successful implementing Dynamics, even at these larger scales, and the business outcomes that they've gotten. 'Cause that's... Like, I think if there's anything that's super clear, is, I mean, we are a challenger, so if you go with Dynamics, we have to deliver, and it can't just be like for like. It has to produce good business outcomes. And I think we've built a very impressive collection of customer stories and references at the enterprise Fortune 500 level where people have used it. And one of the big things is we, of course, think we have the best product at the best price, but there isn't a company in the Fortune 500 that doesn't already have a CRM or ERP. I mean, like, how would you even do business without one? So in all cases, we're trying to find a way where we can get in and win a workload. And if we go to a company and say, "Let's replace the ERP in your HQ, and it's going to cost $1 billion in five years," that's a hard place to get started. Instead, what we've been doing is we're peeling out workflow by workflow into our modular, built for the cloud, primarily cloud-based solution, to basically win that first set of business. And what we see is, you get one workload, you show the results. We think best outcomes at best price. Then you get a second workload, third workload, fourth workload, and instead of trying to win the whole CRM or the whole ERP at one time, we can win a workflow at a time. And that is how we've gotten into the enterprise, and that's how we've been able to grow there, so successfully over the last couple of years, is because it's not a big rip and replace. You can get value in areas where you need it. And kind of if I go layer on top of that, to think about, like, where we go in the future, the workflows are gonna be where you start for AI transformation. Like, your ERP or your CRM, it, it'll be AI powered, but you wouldn't want to migrate your ERP to get AI. You'd instead want to migrate the processes that have the biggest business opportunities. So you want to, say, improve customer support for your consumer division because that's the biggest ROI you can get. Let's go start there, as opposed to trying to replace everything all at once. That makes sense, and it seems like it's working well for you guys. I wanna maybe pivot to something else. When we ask folks in the industry about the killer use case for generative AI, so far, the one that comes up most often is customer service and support. You've rolled out a lot there, including Dynamics 365 Contact Center. What have you seen in terms of customer adoption, and what differentiates Microsoft from numerous others going after this opportunity? Yeah, I mean, I'll, I'll go back to something I mentioned earlier, which is, like, context is kind of everything. Yep. Context, like, that's the data, that's the workflow, and it's also the feedback signal, so as the AI runs, learning, did it work well or did it not, so you can use that to improve, like, ranking and how you select context for the future. Something that we've done in the contact center, which we've been working on for four or five years now to get to this point, is we actually have one of the broadest contexts as it relates to customer service and support. We have Copilot Studio to build things like IVRs and chatbots. We have Dynamics 365 Contact Center to provide the communication channels and engagement channels, so that's voice, that's digital channels, and so on, and we have the agent desktop or case management solution in Dynamics 365 Customer Service. If you go look out at the market, who has a leading conversational experience for AI? Who has a leading engagement channel experience for contact center, and who has case management? There's really not anyone that does that. People have bits and pieces. You know, you got AWS Connect for the channels, but not the front and the back. You got Salesforce for case management, but not, say, the other components built in. So we think we have the comprehensive left-to-right contact center solution with a great bundled price to cover the entire thing, and that gives the AI, which is at the heart of this whole story, an unbelievable amount of context. We can know, did the chatbot solve the problem, to what did the agent get for context, and did it produce a good customer experience? We have a very broad feedback loop, which makes it very compelling. And the place we're going is, we think we can ultimately produce something that is nearly an autonomous contact center. So this idea, it's almost like an appliance which goes from start to back, to run the entire customer support experience. So that's kind of what we've been working on, and we've been able to leverage a lot of great assets we already had at Microsoft, from, like, Teams for the communication channels and, Copilot Studio, which have been built out for other places. So we feel really bullish on this space, and we've used this at Microsoft, and I would say it's probably the most impactful GenAI project that we've used inside of Microsoft already to provide our own customer support across something like forty thousand agents and ninety contact centers. We've seen pretty incredible uplifts. We have a great white paper that we published about how we measured and experimented on the business impact. But, like, this is a place where we think there's real, real opportunity for both consumer and commercial companies today to use AI. Got it. I, with respect to time, I've got one more that I'm gonna try to condense into one last quick question. But, you know, when we think about Copilot, everything you've said about context, interoperability, Common Data Model across Azure and Microsoft apps, and all those integrations, Salesforce, ServiceNow, Adobe, what have you, what makes Copilot best positioned to be the point of integration across all these systems? I mean, I think you've spoken to this a bit already, but does extensibility ultimately expand the TAM for Copilot? And is this scenario where it becomes the orchestration and engagement layer for other Copilots and apps? I mean, I know in software, it's always been about coopetition and a win-win to, you know, sort of broaden the pie for everybody else, but we'd love your thoughts on that. Yeah, like, I mean, the first thing is, I think with GenAI, like, the overall software market is gonna get a lot bigger. 'Cause if you produce these business outcomes, there's a whole lot of investment that can happen. So there's a lot of room to partner and do coopetition that works best for everybody. If we look at, like, Microsoft's unique position, kinda where is the canonical workspace and workflow for a typical office worker? It's Microsoft. It's Windows, Microsoft 365, Office 365, it's Outlook, Teams, Excel, PowerPoint. That's where people start their day and run most of the business process each and every day. So we think that is a really unique position for us, and we're an open platform, we're an open company. We have partnerships that we've already announced with SAP, where their Joule, their AI assistant, will integrate inside of Microsoft Copilot. We have partnerships with ServiceNow, where Now Assist will integrate inside of Microsoft Copilot. We really think that you can start with Microsoft Copilot as the place you begin your day at work, and from there, it will extend out to other solutions and other AI components. And we are very committed to make this a platform, almost like an operating system for like the GenAI assistants and agents you see at work, that our customers can use. And, so that, that's our big focus. That's why we've been so attentive to platform and extensibility for Microsoft Copilot very early on, and while we've gone out and done a bunch of these different partnerships with other companies, whether they're industry or SaaS providers, to extend Microsoft Copilot. Awesome. Charles, with that, we are out of time. Thank you so much. It's amazing to see you out here at the Deutsche Bank Tech Conference. Thanks, Brad. Thanks, everybody.