Salesforce, Inc. (CRM) Earnings Call Transcript & Summary

Hello, everyone, and welcome. My name is Teri Hatfield and I am the Executive Vice President of Tableau Global Sales. I'm here in our New York office, super excited for this session today. I think it's going to be great, and I can't wait to hear from all of you. There's a 1,000-plus of you here today for this virtual workshop, and we're about to dive into Tableau's latest AI innovations and trust me, no matter what your technical background, there's something for everybody here. So today is all about showing you how AI can make a real impact on your business and get some really solid outcomes right out of the box. So in this presentation, we'll be taking a look at some forward-looking content from the Tableau team. So I'll remind you that all customers should base any purchasing decisions on currently available products, and I'm sure you read all of that. So you're all fast readers, I just know it. All right. But in all, I really want to say thank you so much for being here with us today. We're so excited to go on this ever fast-changing journey together, and these are the types of sessions that really help us continue to advance our work together, which is fantastic. So before we dive into the details, let's start with a quick ice breaker to get to know each other a little bit better. I'd love for you to jump into the chat and share your thoughts on a couple of questions. So what is one AI-related task you wish you could automate right now? All right. Documentation, billing scheduling, report generation, onboarding. I love these. Updating Jira, upgrading documentation. Really the onboarding thing. I think that's something that could make a very, very increased amount of productivity. Attending webinars. Maybe in the end, you won't say that. I don't know. Insights, data QA. All right. I love these answers. It's amazing to see how just creative and diverse the community is and just all the different things that everyone is thinking about. It's exactly the kind of energy we're going to need for today's workshop, which is great. ETL process. Okay, keep going. All right. So you probably want to know what our plan is for today. We're going to kick things off by exploring Tableau Agent where we'll use natural language to analyze service metrics. Then we're going to move into a hands-on exercise where you'll get to create your very own AI-powered agent. Finally, we'll wrap up by looking at how to measure the impact of your new Tableau Agent using Tableau Pulse. So by the end of the session, you'll walk away with practical skills, hopefully, that you can immediately start applying in your business. You should have received an e-mail. So hopefully, you'll have that ahead of today's workshop asking you to register for Trailhead and to set up your Salesforce, your Agentforce Playground. So if you haven't, Jeremy, who I was going to introduce in a moment, is going to walk you through how to track this down. So speaking of Jeremy, I'm excited to introduce the lineup of speakers who are going to guide us through our journey today. First, we have Jeremy Blaney, Senior Director of Tableau Product Marketing. He's going to walk us through how AI is transforming the analytics landscape and give you a behind the scenes look at our latest features. Next up will be Kiyoshi Jones, who I have known for quite a long time. Senior Director of Tableau Solutions Engineering. He's fantastic. And he is going to demonstrate the power of Tableau Agent and Pulse. And finally, Shauna Goldman, Tableau Product Marketing Manager. She's going to lead us in the hands-on session and make sure you're equipped to build and deploy your agent successfully. We are so lucky to have this incredible team with us, and I can tell you firsthand, I'm 13 years in, and I've known them all for quite a long time. And before I pass the mic off to Jeremy, I encourage you to participate in the chat, ask questions and share your thoughts throughout the workshop. The more you engage, the more you're going to get out of the session. And obviously, the more we also learn about you and things that we can do to help you out. So if you run into any roadblocks, don't worry, just drop your questions in the Q&A box and our solutions engineers will help you as soon as they can. They are rock stars. I just saw that comment. That's exactly right.

All right. And over to me. Hello, everybody. It is so great to meet with you today.

I'm sorry, Jeremy. Over to you.

It's all good. It's all good. Hey, we do it live. Hey, everybody. Good morning, good afternoon. It is wonderful to be with you all today. Some of you may have done the prework. Some of you may not have had a chance to do. If you're in that latter group, no worries, let's take a minute just to get you set up for today's workshop. I'm going to walk through this kind of fast, but all signs point to a specific spot where you can get more guidance. So first, I want you to do this. I want you to open the Configure an Agentforce Service Agent Trail and Build Your First Agent step-by-step guide. You can find this linked in the resources box on your screen. In step 1 of that guide, what you need to do is click the link to the Configure an Agentforce Service Agent Trail, excuse me. At the top of that page, there's a little box that says this badge requires a new custom Agentforce Playground. From there, I want you to click, get started for free. You'll be prompted to sign in, if you don't have an account, take a moment to create one. Once you're signed in, then you will head back to the trail page. There, you should now see the options to either connect to your playground or launch it. When you do this, you will also get an e-mail, as was just said, from Salesforce with a link to launch your playground. When you open it, you'll see the Coral Cloud Resorts environment. This is where we are going to spend the majority of our time today. Again, I know we walked through that kind of fast, don't worry. You have about 10 minutes or so before your playground needs to be up and running before we actually put our hands in the soil and get started building that agent. So let's start here just to set the stage of why we are here today. AI is obviously the talk of the town. And it's not because it's new, but because we're stepping into this new era. But let's be honest, reality for those of us in the business of data and analytics is that we are still wrestling with fundamental challenges. Let's talk about analysts. Analysts are stretched thin, all of you are constantly juggling a million different things. I've yet to meet an analyst who says, you know what, I finish everything in 40 hours a week. It's almost always the opposite. I can't get it all done. I'm buried in data prep. I'm building just a ton of dashboards. I'm wasting a bunch of time reconciling inconsistencies and so on and so forth. And it's not just analysts who are struggling, let's talk about the business. Business users have their own hurdles to overcome, finding insights can feel like something of a treasure hunt, right? Dashboards are everywhere and sifting through dozens of pieces of analytics content just to start your day is a far too common sort of task or occurrence. And then there's the data itself. The landscape is large and fragmented, silos persist and too many business tools that many of us use as a part of our jobs, they lack real analytical depth. And as a result of that, we're forced to switch between systems, just to answer very basic questions that give us a sense of how now to act with regards to achieving our business goals. AI and agents in particular hold the potential to solve these challenges and transform not just how we are working with data, but how work itself changes. But let's be honest, getting started with AI is easier said than done. Well, that's no longer the case, thanks to Agentforce. Agentforce is the agentic layer of the Salesforce platform. It is designed to deploy autonomous AI agents across any business function. It gives you the tools to create and customize agents that work side-by-side with your teams. These agents can handle all kinds of tasks autonomously or assistively. One standout tool or capability in Agentforce is a library of ready-to-use skills. You can think of skills as bundles of agentic capability, if you will, that are built for a range of use cases. Now those use cases could include lead development for sales. These use cases could include recruiting if you are in the world of HR, it could include customer service for all service. And yes, it could also even include data and analytics. And that's what I want to focus in on just for a few minutes. So let's dive into those prebuilt skills within Agentforce that are purposely built for data and analytics. So DataPro. This is the first skill. It is your partner for data prep, data management, data visualization and analysis. It works both alongside you and autonomously taking on jobs so that way you can move faster and spend more time on strategic analysis, which is an ask that I hear of every analyst that I talk with. It doesn't matter who I'm talking to, they all say the same thing. I want to spend more time on strategic analysis. Concierge really does revolutionize how you interact with your data, it gives you the ability to ask questions of your data and get answers through natural language Q&A. I am especially excited about this because it unlocks a completely new way of working with data. It lowers the barrier to analytics for everyone. Well, for years, people have long wanted the ability to chat with their data, that need, that one, that desire is not new. But let's be honest, like those experiences haven't exactly lived up to expectations. Well, that's no longer the case. Thanks to LLMs and agents. Agents are taking on the heavy lifting here. They process those complex natural language queries. They map those queries to data. They adapt to your unique business context through semantic understanding. And then they generate that accurate context-rich response, which increases trust in AI as a whole. Let's talk about Inspector. Inspector helps you stay focused on what matters most by proactively surfacing key insights. It does this by basically monitoring your metrics at scale. It's constantly evaluating what's happening and why this is going to be huge for identifying business opportunities that are hidden within our data. And then finally, we have analytics apps. This skill makes it super easy to create dashboards using templates and such. You can also use the skill to embed actionability directly within those dashboards, integrating them with other business tools and systems like Workday, for example, so instead of ending up with a static dashboard, which has great value, don't get me wrong, you get this super dashboard. You get this dynamic app that lets you take action both seamlessly and right away. Now these skills are natively integrated into Tableau, and that includes the new Tableau, which is formerly known as Tableau Einstein and Tableau Cloud. Today, we're going to take a deep dive into how you can leverage the DataPro, Concierge and Inspector skills directly in Tableau Cloud. We're going to show you how these skills empower everyone, from analysts to business users to work faster with data, to uncover problems that need solving and to take some sort of action to solve that particular problem. Now you can also pair these skills with Agentforce to enrich every agent with analytical capabilities. In the coming months, you're going to hear more about this. But today, we're going to zero in on the agent building experience. So how do you build an agent and empower it with skills made up of topics and actions? We're going to walk you through this step-by-step using your demo environment as the foundation. Our goal in all of this is really twofold. Like one, we want to show you how agents solve business problems and deliver incredible value for your organization. And then two, we want to spark your imagination, right, for how agents can enhance data and analytics in exciting, new and impactful ways. Yes, you can use these skills directly in Tableau through native integration, as I said previously. But what makes all of this even more powerful is the ability to create agents that have analytic skills. What does that mean for you? Well, it means that you can build agents for internal use, like sales or HR that feature analytical capabilities or you can create these client-facing agents that deliver more than just like plain text responses. They can share dynamic data visualizations. They can pull from your semantic data models to ensure the response is accurate and a whole lot more. I know that all of this is a little meta, and may be difficult to like fully grok. Well, I believe that the best way to learn is by doing, and that is why we are here today. I am very excited about the future of analytics and now is the time to create it, I'm here with you and I am thrilled to have you along for this journey. Without further ado, let's get started.

Excellent. Thank you, Jeremy, and good morning, data fan. Thank you for joining us. I'm excited to kick off the first part of the workshop, which will start with the demonstration of Tableau Agent. Reminder, Tableau Agent is available and powering analytical experiences today. So if you haven't tried it out already, turn on Tableau Agent, try this out for yourself. Now as I share my screen, let me give you some brief context to the theme surrounding today's exercise. You heard it just a moment ago from Jeremy, but we are going to be working with a fictitious resort called Coral Cloud Resorts, which is going to be the basis upon which we'll be exploring agentic capabilities. Coral Cloud Resorts is looking to use data and AI to deliver highly personalized guest experiences. Here we are in Tableau Cloud, where we're going to start by connecting the data that we have in Data Cloud starting in workbook. Now just to orient everyone to what we're seeing here, we are connecting to some data that I have in Data Cloud that contains information about service cases. On the left-hand side, we can see that the service case information includes things like agent, which agent perhaps answer that case for that call. The customers that are impacted, the departments of which these cases are rolling into and then metrics such as CSAT for customer satisfaction, talk time and wait time. Now I've heard that there have been some challenges in the recent increase in the amount of customer cases that we've had, so I'm going to use Tableau Agent to help me out here. Let's ask the question, show me the count of cases by day or the last 3 months. Wow, look at that, it's created a visualization so that I can see and confirm that, in fact, we're having an increase of cases. And when I specified 3 months, it looks like it actually put a filter on and ensure enough, this trend is unsettling. We went from having about 400 cases, just a few weeks back, and now we're seeing an enormous spike here. All right, maybe it was showing a little bit further. Break this down by department. All right. I'm using natural language with Tableau Agent, and Tableau Agent is returning visualizations for me. Once again, I'm seeing this broken down by department just as I asked that the visual indicator that's popping out to me is this anomalous reservations department. This green line keeps going up quite higher. So I think these cases are probably having something to do with the reservations. So maybe let's ask a few different questions. Now with Tableau Agent, we can keep iterating on the sheet. I'm going to create a new sheet, so I can ask some more questions. Let's see how the wait times have been trending. Show me average CSAT by month. Once again, CSAT, for those that aren't familiar, is customer satisfaction. I can see that customer satisfaction is going down. Cases are going up. All right. How are we doing with the wait time? Remember, we have that wait time in here, let's ask. Show me average wait time by month. All right. Average wait time went from about 30 seconds here. Look, 217 seconds. Yikes. Customer satisfaction is definitely taking a hit. Our wait times are going up as we see here, customer satisfaction is going down in the same period. Okay. I'm going to iterate a little bit here. Let's break this one down by department. There's those reservations again. This green line here, definitely something going on there. So let's take a look at some questions so far. I know that I've gone somewhat quickly, but I did want to demonstrate how you can use Tableau Agent to be able to ask questions in natural language and get that visual feedback. So I'm curious to see what folks think so far, I see a couple of questions coming through. What if you don't know what questions to ask to begin with? It's a great question, sort of that blank canvas problem, if you will. Well, let me start with a new sheet here and based off of the data that I've connected to, Tableau Agent can actually provide suggestions or ways that we can get started. So tells me build these, filter and sort or I might say, what do you suggest Tableau Agent, where might I dig in? It gives me some examples of some prompts or questions that I can ask. Here's another question coming in. I'm presuming Tableau Agent was formally called Tableau Einstein Copilot. Yes, Tableau Agent is an enhancement and iteration of what you might have seen with Copilot previously. So in case you're curious, that is accurate. All right. Can I create a calculation to convert to minutes? Okay. All right. I see you data fan, throwing a curveball here. Let's see. Create a calculation to convert wait time to minutes. Now we are dealing with Tableau Agent, which goes through a trust layer, and it is using a large language model on the back end. So it is interpreting this information in a trusted way and then providing a response for us. Wait time in minutes, we take the wait time in seconds and divide it by 60. Not too bad. If there's any other questions. Can it apply filters? Yes, we saw an example of that earlier, but let me just show real quickly here, show me a time series of CSAT. I'll just get visualization again. Okay. And then from here, I could say filter to reservations. All right. I think that's actually pretty powerful with what's happening here. I use natural language once again, filtered to reservations. And if you're having trouble using agent or maybe with inaccurate responses, I'd give it as much detail as possible. But here I'm saying filter to reservations. I actually don't have a dimension here or a field called reservations, but it's able to interpret the question and it found reservations within the department and [ so knows ] to filter that. I think that's pretty powerful. All right. Can you continue to modify the view? Maybe I need a little bit more detail on that. Change what we're seeing. Let's give it a try. So change this view to -- I don't know, talk time instead. Very good. So here, you saw an example. It's building on that iteration. So if you want to do ad-hoc analysis, exploration, create new sheet, here, we're iterating, we're building on this. And in fact, what we're seeing here, once again, to remind everyone, the paradigm or the difference of what we're doing, the operations here, we're using natural language. We have insights. We're domain experts. We know about cases. We need to drill in and we need some visual confirmation for what we're seeing. So instead of having to write SQL or code or script or anything like that, here, we're using natural language. So I'm seeing a lot of questions here, which is fantastic. I love the engagement. So thank you. I'm going to go ahead and stop sharing now and do a brief recap. All right. So that was a look at Tableau Agent. And I think, once again, very exciting. To recap what we just did, we used AI and conversational analytics on trusted data, by the way, and we were able to validate that, one, our customer service case following that Coral Cloud Resorts is increasing, particularly in cases having to do with reservations. We also found that wait times for customers are going up and customer satisfaction during that same time period is going down. That's a big problem. Now I have those insights, not necessarily the insights I wanted, but those are the insights that I got from Tableau Agent. So thank you, Tableau Agent, helping me see and understand my data. Now as I hand it over to Shauna, we're going to transition from Tableau Agent powered by Agentforce and get hands on to see how Agentforce can be used to help improve the experience for our Coral Cloud customers. Over to you, Shauna.

Awesome. Thank you, Kiyoshi. So now we're going to get hands on. And together, we're going to build this customized agent to solve for those service challenges that Kiyoshi shared that Coral Cloud is facing. So again, we saw that spike in customer cases and our customer service team just can't keep up with this demand. So this is where we can create a customer service agent using Agentforce to help solve all these challenges. So what you're going to do is take a look in that resources section. You're going to see a document titled Build Your First Agent Guide. So please open that up because this is going to act as another resource for you during the stage of the workshop. This stage is also going to involve some copy pasting of text for this purpose, just to save you all some time since there's a lot of text to type out. So I'll call it out of what step that's on. It's going to be highlighted and bolded to make it easy to find. But definitely get that PDF open so you can make this easy on yourself. All right. So I'm going to go ahead and share my screen. Awesome. All right. So you should be seeing this Coral Cloud Resorts environment. This is what Jeremy walked us through earlier. So if anyone is having any issues logging in, please reach out to our team. We've got a whole group of us ready to answer any questions, and if you're having any challenges getting in. Just as a reminder, to get to the screen, you went to the Trailhead, connected to the Playground with this button, sometimes it says launch, it can say different things depending where you are. And then once you go through that process, you'll be able to open up the screen. Sometimes it's easier to use the link that Salesforce emails you, so you can try that as well. So once you log in, this is the environment you're going to see. And we have some tactical setup to do before we can really jump into building our agents. So the first thing we're going to do is enable our agents. So you can follow on with me. So what we're going to do to start is in the upper right corner of the screen is this little gear icon. So just click on this gear, it's a Setup icon, so then go to Setup from the drop-down. This will open up in new tab. And then in the upper left, there is the search bar that says Quick Find in it. So you are going to in here, type out Einstein Setup. And this will bring us to Einstein Setup in the drop-down and just backing up a bit, maybe wondering what Quick Find search box, like what are we doing here. This is just a tool to help you go where you want to quickly. So you can enter a search term into this box, and then this is going to filter the list of pages on this environment so then you can just find what you're looking for faster than you would by browsing through the site. So I'm going to select Einstein Setup from this drop-down and this will bring us to this new page where we can turn Einstein on. So there's a little grayed-out toggle. So you're going to just click it. And once you press it to be on, it's going to turn green and you'll know you're ready to go. And then back in this Quick Find search bar, we're going to search and select Agents. So select Agents from that drop-down and then under Get to Know Agent, there's a section that says Agentforce. So it's currently toggled off. So just press on the toggle and you're going to turn this Agentforce to be on. All right. So next, we're going to publish the Coral Cloud Experience Cloud site. So this is essentially a design view of Coral Cloud's customer-facing site. So we're just going to set this up now. So later, we can add the agent to Coral Cloud's customer-facing site. So back in this search bar, we're going to search and select All Sites. So pick All Sites from that drop-down. And then this will bring us to a page where we have 2 options here. So the second option is Coral Cloud. So to the left of Coral Cloud, you'll see this link that says Builder. So that's what you're going to want to click on. And this will open up a new page that's going to be that kind of design view of Coral Cloud's customer website. If you get this pop-up, just press okay. And then you don't have to worry about any of these photos or blocks here. All we're going to do is publish the site so we can access it later. So in the upper right corner, you're going to press Publish, use this button. And then when it to asks you to confirm, do you want to publish your site? Yes, we're going to publish this. And then once this loads, we're going to press just to confirm that we get it. All right. Got it. And Salesforce will likely e-mail you just a confirmation that you published to your site. So if you get an e-mail from Salesforce about publishing a site, that's what it's about. So in the upper left corner of the screen, you're going to see these 3 kinds of rectangle boxes that form a square, press that, and then from the drop-down select Salesforce Setup, and this will just bring you back to the Coral Cloud Resort environment, so we can continue to work on the agents. All right. So the last piece of setup that we need to do here is we need to add in some permissions for this agent. So back in that Quick Find search bar, we're going to search and select Users. So select Users from this drop-down. Oh, it didn't work. All right. So then we have this whole list of users. And the only one that we want to select here is the third from the top, it says EinsteinServiceAgent User as the full name. So we're just going to click on that name for EinsteinServiceAgent User. And this will bring us to that user page for EinsteinServiceAgent. And we have all these details. We don't need to worry about any of this. We're going to scroll right past it and there's this permission set assignments block. So here, there's a button to the right of permission set assignments that just says add assignments. So you're going to click on this button so we can add it some assignments. All right. This will bring us a new page. On the left, you'll see all of the available permission sets that we can add to this agent. And on the right, you'll see what's currently enabled. So we have -- you should see Agentforce Service Agent user as an enabled permission set. We want to add 1 more here. So you're going to scroll down on this left bar to service agent permissions. Here we go. Okay. So you'll see service agent permissions here, so I'm just going to select it and then click add. One thing to be careful of, there is a service agent permission set. Make sure that you're clicking service agent permissions. If you get an error message that something went wrong, you might have selected service agent. So I've done that before. So just make sure that you're selecting the correct permission set here. And then once you have these 2 permission sets enabled, click save. All right. And then we've done all of our tactical setup. So I'm just going to pause briefly here. How are you guys doing? This is the tactical setup to build the agent, but now we're going to get into the actual build. So if you have any questions, be sure to ask our team, we're ready to help you and you also have your guide to help you with some screenshots here. All right. So now we're going to go ahead and create our agents. So in that Quick Find search bar, you're going to search and select agents. All right. And then we've been on this page before to turn on Agentforce, but now in the upper right corner of the screen, you should see this button for plus New Agent. So this is how we begin building this new agent. So go ahead and click this button, and then now we're going to select the type of agent we want to have. So there's only 1 type here, so just select Agentforce Service Agent, and then hit next. All right. Here is where we're going to review these prebuilt topics. We're going to create our own custom topic today. So going to deselect all of these topics except for the very last one, which is General FAQ. So to deselect it, you're just going to click this button that says checkmark added on the right-hand side of these blocks. So I'm going to deselect each of these topics and then leave the General FAQ as an added topic. All right. So once you do that, there is a button click Next in the upper right, and then here, we're going to define some settings for this agent. So we want to name this agent something more unique to Coral Cloud. So we're going to call it CC_Service_Agent. As you can see from this drop-down, I've made a lot of agents here. So you want to make sure the API name also changes to CC_Service_Agent. So just confirm that that's changing with you -- for you as well. And then description and role, there's already some text populated, so you can leave this as it is. But for a company, we're going to want to copy paste this from the guide. So if you scroll down in the guide to step 5, this gives an overview of what Coral Cloud is. You can just copy paste your text in, that's what I'm going to do. And I know that if you copy paste from the guide, your text might have some line breaks, but don't worry if that happens because it's still going to work just fine. So we're sharing that Coral Cloud Resorts provides customers with use exceptional destination activities, unforgettable experiences and reservation services and, of course, top-notch customer service. For agent user, we're going to select the EinsteinServiceAgent User that we added those permission sets to before. And then that's all we need to do on this page. So we updated the name, added some text for company and then updated agent user. And then we'll click Next. All right. So we're not going to use Data Cloud for this workshop, so we can skip right past this and click Create. All right. So it looks like I got some errors. So if you've got some errors, no worries, I'm not going to worry about these, so you don't need to either. In the left-hand side, there is the topic details section. We're next going to add some custom topics and actions for this agent. So if you think about what a topic is, it's essentially a job to be done for our agent. So remembering our analysis with Tableau Agent, we saw that Coral Cloud's customer service department has been receiving a big uptick in cases, especially with making reservations for those activities. So this is something that we want our agents to be able to help with when having conversations with our customers. So this is something we should definitely keep in mind when creating this topic or this kind of job for our agent to do. So in this Agent Builder, we're going to click the New for topic details and select new topic. All right. So this will open up a new window for creating a topic. So here we can again populate some fields for this agent. So if you want to copy paste pull up step 6 in the guide. So you can fill in these text box a little easier. So the topic label, this is what represents the job to be done within the topic. So here we're going to title this Experience Management. And then classification description. This is what guides the LLM on when to select this topic. So here, we're going to give us some details about how this is going to address some customer questions and any issues that we have with making reservations. So you can copy paste this text from the guide. All right. So this text shares how this topic addresses customer inquiries and issues with booking experiences at the resort. And then for scope, this is what determines what the agent can do in this topic. So it will constrain the agent not to respond outside of the given scope. So here, we can enter some info about making sure this is for reservations and experiences that are specifically for Coral Cloud. So I'm going to copy and paste this text from the scope. So shares the agent's job is to assist users in navigating and managing bookings for different experiences, making sure it's a seamless experience. All right. Then if you scroll down, there's a section for instructions. So the instructions are what guide the LLM on how to best use actions to perform the topic's job. So you'll see there's a button to add instruction. So we can continue to add more instructions as we go through this process. But for now, we're just going to add in 1 set of instructions to get us going. So this will be about running an action to get customers some more information on activities. So I'm going to copy and paste this first instruction into the block. So this shares if a customer would like more information on activities or experiences, you run this Get Experience Details action. All right. So once you paste in all of this different text for topic label, classification description, scope, instructions, we can then hit Next. And here are some pre-built actions, we're not going to use any of these pre-built actions. We're going to create our own customized actions for this workshop. So you don't need to worry about this. You can just click Finish. All right. And that will have helped us to build our very first topic. So you'll see topic label Experience Management will populate under here. So now that we have this topic, we're going to create some custom actions to go with it. So this topic is the job to be done, you can think of actions as the tools to actually do this job. So actions are added to the topic to limit the agents, so they're only using the relevant tools for the given job to be done. So the first action we're going to add is about getting experience details. So this makes sure that the agent can share details about each experience a customer might be interested in. So for instance, if I want to learn more about an experience like a kayaking trip that Coral Cloud had, I'd be able to ask the agent and the agent will be able to share more information about what that experience looks like. So to create an action, we're currently in this topic details page. So we're going to click into that topic we just created. So click the link for Experience Management. And now I'll bring you to this topic configuration page. So this is all the text we just entered to create the topic. So we're going to leave this as is. But if you go to the right to This Topic's Actions sub-tab, I'm going to click on this tab. And under here, you'll see a button that says New and then Add Action. So this will enable us to start adding new actions specifically to this topic. So what we're going to do is for Reference Action Type, we're going to select Flow and for Reference Action, we're going to select Get Experience Details. So just backing up a little bit, might be wondering, what are we doing here, what is the reference action. So this is a predefined action that can be used as part of a custom agent action. And you'll see there are 3 different types of reference actions. There's apex, flow and prompt template. For this action, I'm referencing a flow. So flows implement some deterministic logic. So this agent has this predefined step-by-step process to go through. And when we leverage these Salesforce flows, the agent has a more structured path to reliably follow and integrate more seamlessly with tools. And it can also execute business logic consistently. So all of this ensures that we're getting repeatable and reliable outcomes when using this reference action. All right. So you'll see that Agent Action Label populated as Get Experience Details, same with the API name. So you can leave this as is and then click Next. All right. So here, we're going to just check some boxes under Input and Output sections. So under experienceName, we're going to check Require input on the left hand-side. And on the right-hand side for the green box for Output, we're going to check Show in conversation for experienceRecord. And then we can click Finish, and we'll have just created our first action. So you'll see it populate under this Agent Action Label. So now we're going to create another action to validate customer details. So for security purposes, we really want our agent to make sure that whenever they're talking to a customer, they're saying -- making sure that they can confirm who the customer is with some key contact details like asking that customer for their e-mail or member ID number at the start of any conversation before that agent gives this customer more info about Coral Cloud. So we're still going to be in This Topic's Actions sub-tab. We're going to repeat some sub-tabs that we just did, we are going to go to New and then Add Action with this big plus mark. And for Reference Action Type, this is also going to be a Flow. And then for Reference Action, we're going to just check Get Customer Details. And then you can leave these labels as they are and then click Next. All right. Here we are going to check some more boxes under Input and Output. So on the left-hand side, under e-mail inputs, we're going to check Require input and still under Input, we're going to check Require input for memberNumber? And then on the right, under Output, that green box, we're going to check Show in conversation for the contact. And then you can click Finish. All right. So that is our second action created. So it would also be helpful for our agent to get the available session records for each of the experiences at the resort. So we're going to create another action to do just that. So still on This Topic's Actions sub-tab, we're going to click New and then Add Action. So this Reference Action Type is also going to be a Flow. And for Reference Action, we're going to be selecting Get Sessions. So select Get Sessions, and you can leave these labels again as they are, so we'll name our action. All right. And then under Input, for experienceid, you're going to check Require input and for startDate, you're going to check Require input. And under Output for sessions, you're going to check Show in conversation. And then you can click Finish. All right. So I think these customers are also going to appreciate having a more personalized touch. So let's create an action to give more tailored recommendations to our customers and making sure that we're taking their current schedule into account when talking to them. So for instance, if I have booked jet skiing as an experience for tomorrow, when I'm asking the agent for more recommendations about activities for today, I don't want that agent to be recommending that I go jet skiing again. So this action is going to help make sure that these agents are tailoring their recommendations. So go to New, Add Action. And for Reference Action Type, this time, we're going to select Prompt Template. So we've only been using Flow so far. So this time, we're doing a different Reference Action Type for Prompt Template. So what does a prompt template mean, this is going to guide the LLM to generate contextual dynamic responses. So by grounding templates in relevant data, knowledge bases or kind of specific user context, prompt templates can ensure that the agent is providing accurate data-driven and personalized outputs. So this Reference Action Type is going to be ideal for conversational adaptability and also leveraging existing information. So that's exactly what we're going to want to -- for this tailoring to the personal schedules for our customers. So for Reference Action, this time, there's only one Prompt Template reference action. So that makes it easy on us, just pick Generate Personalized Schedule. And then for Agent Action, you can leave these labels as they are and then click Next. All right. So this stage of the action creation for personalized schedule actually has some instructions to add. So if you pull up your guide and scroll down to step 10, we're going to copy paste some instructions into these text boxes. So under Agent Action Instructions at the very top, we're going to enter some instructions about creating a personalized schedule with the available experiences that match the guest interests. So again, this is at step 10 in your guide. So I'm going to copy and paste some text in. So this just shares we're going to generate a personalized schedule that includes the time and location of the resort experiences that are available today and match those guests' interests. And then we have one other set of instructions to copy and paste in. So under Input on that left-hand side for Contact, we're going to add in some instructions. I'm going to copy and paste. All right. So these are going to be details about how this should represent the contact info and how this is chained from executing our Get Customer Details action. And then for Prompt Response, we're going to check Show in conversation. All right. So again, we added instructions for agent action instructions and then also for the contact. And the only other thing we have to do on this page is check Show in conversation under Prompt Response. And then we can click Finish and we finished our personalized schedule action. All right. So we have one more action to add. So this is to create a booking. So if a customer wants to book one of these experiences that the agent recommended, we're going to give the agent the ability to quickly make that reservation. So for instance, if you're talking to this agent and they share this recommendation for jet skiing, for instance, and you decide you want to book that experience, you don't have to go anywhere else. You can just continue talking to the agent and write in the chat, be able to make that reservation. So we're going to go to New and then Add Action. So this Reference Action Type is also going to be a Flow. So we're going to go back to the flow reference actions. And for Reference Action, we're going to select Create Experience Session Booking. And then you can leave these labels as they are. All right. And then for the Input section, we've got a few boxes to check more so than we have in the past. So in the Input under Contact_Id, we're going to check Require input. And then for Guests, check both Require input and Collect data from user. And for Session_Id, check Require input. And then on the Output section, we're going to check Show in conversation for Booking. And for Output_Message, again, you're going to check Show in conversation. So just to reiterate, since we had more boxes to check than we have in the past, under Input, we're checking Require input for Contact_Id. For Guests, we're checking both Require input and Collect data from user. Session_Id, we're checking Require input. And then under Output, we're checking Show in conversation for Booking and Show in conversation for Output_Message. And then we can hit finish. Thinking about it. Okay. So we have our 5 actions now once this loads in. Here you go. I'm going to assume it's going to load. I'm not going to worry about that. So once we -- now that we have our 5 actions added to this agent, we're going to be able to add to our instructions so we can effectively guide this agent and how to best use these actions to perform its job. Yes, look. It loaded. It was just too early. So what we're going to do is we're on This Topic's Actions sub-tab. So we're going to go back to Topic Configuration under Topic Details. So we're in the Topic Configuration sub-tab. And again, here's all the text that we added when we first created this topic. So at the bottom, you see the Section 4 instructions. So we have this very first instruction that we added in and then there's this button to add instructions. So we're going to click on this button to add instructions. Click on this 5 times to create 5 blank instruction boxes. All right. So -- and here, we're going to add instructions for each of the actions that we created. So this is just going to be a natural language instruction to guide this agent and how to best use these actions we created to perform the job that we want this agent to do. So as you can imagine, we're going to copy paste some instructions here. So pull out the guide and scroll down to step 12, and you'll be able to copy and paste into these instruction boxes. So we're going to leave this first instruction box as it is. We don't need to worry about changing this. We can keep this from when we created the agent, we created the topic originally. We're going to start in the second instruction box here. And what we're going to do is we're going to copy paste to tell the agent to ask for specific customer details at the start of every conversation. All right. So if the customer is not known, they're going to always ask for that e-mail address and that member number in order to get the contact record. And then for the third instruction box, we're going to make sure that the agent is communicating with customers about experiences that haven't taken place just yet. I'm going to copy and paste. So this instruction shares that whenever a date is provided, we're converting this date to the specific format and that we're making sure that this date has not occurred in the past that all of these experiences are going to be forward-looking experiences. And then in the fourth instruction box, we're going to give the agents some details on how to respond to questions about sessions. I'm going to copy and paste this fourth instruction. All right. So if the agents ask to get sessions for the experience, they're using that Get Sessions action. They're asking for the day, they may also use that Get Experience Details action. All right. And then in the fifth instruction box, we're going to provide some info about how this agent should effectively book a session. So I'm going to copy and paste. All right. So if the agents ask to book a session, they're going to use that create booking action as well as Get Sessions in order to effectively go through that action and create an experience booking. And then in the final instruction box, we're going to give some additional guidance to provide recommendations for each individual customer. So I'm going to copy and paste this final set of instructions. This shares if the agents ask to recommend experiences that the user is going to be interested in, they're going to use that Generate Personalized Schedule action to create a schedule that's best based on that customer's interest. All right. So we have filled out 6 instruction boxes now for -- a total of 6, we filled out 5 new ones. So you should have all of these filled out as set a part of your new topic, of your Experience Management topic. So once you have these instructions added, you can click Save at the bottom. My screen is a little cut off, but you can click Save and this will save your topic. All right. So now that we have added these actions to our topic, we've added instructions to our topic, we're going to be ready to test this agent. So I'm just going to pause if anyone -- I know we're going pretty fast here. But if anyone has any questions, please feel free to continue to ask our team, and we'll try to continue to help you. You also have the guide to reference throughout this process. All right. So once you have your topic and actions in this -- in Experience Management in the upper right corner, you'll see a button to activate our agent. So click that button and then we're going to be ready to try this out. So taking a bit of a step back, we've only been working in this one part of the screen, this topic section. But this whole screen, this is Agent Builder, right? So on the left-hand side is topics. This is where we created the topic. We can add actions to our topic and continue to work on editing in that, and that's what really goes into the agents. On the right-hand side is the conversation preview. So this is where we can chat with the agent, act as a customer of Coral Cloud and validate the agent is responding to prompts as we expect it to. And then in the middle of the screen, this is a really cool part of Agent Builder where you can see how the agent is responding to our prompts and how it works with those topics and actions that we created to create an effective response to our prompt and how the reasoning works and the kind of behind the scenes of this agent. So now we're going to put our work to the test and try communicating with this agent. So I'm going to refresh this conversation preview, just to make sure we're starting from a fresh clean slate. So we have some prompts in step 13 of the guide that you can copy paste. You can also go a bit off script if you want and enter whatever prompts you would like to as a customer of Coral Cloud. So what I'm going to do first is I'm just going to ask for more information about one of our experiences. So I'm going to ask for more info about this full moon beach party that Coral Cloud runs. All right. So the agent responds asking for my e-mail and member number. So this is a great response. But one thing I'm noticing in the very middle of the screen under Experience Management, this is our topic. We see our 6 instructions. But for actions, this has zero, and I know we created 5 actions. So this is going to create some errors and issues for me if I continue to talk with this agent. So I'm going to deactivate this agent. And I imagine if I'm seeing this zero action, some of you might be as well. So I'm going to deactivate the agent, click okay. And this is all good, no worries, we can fix this quickly. Under topic label, we're going to click into our Experience Management topic. And we are on the Topic Configuration page. So go over to This Topic's Actions. Yes. And as you can see, there's zero items here. So we're missing our 5 actions that we just created. So no worries. We do not have to go through that whole process of creating the 5 actions again from scratch. We can add them from our library. So if you go to New and then there's a little globe icon and Add from Asset Library, we're going to click this from the drop-down. All right. So here, we're going to search for the actions that we want to include. So it's going to be a test of our memory to make sure that we're getting all of the actions back in. So I'm going to start with Get Customer Details. We also had Get Experience Details. We also had Get Sessions, Personalized Schedule and then one more, we want to create bookings. All right. So I know I did that quickly. So just to find the actions that you want to include, you're going to, in this search bar, start typing out the names of those actions that we created, and you'll be able to find the Get Sessions, Get Experience Details, Get Customer Details, Generate Personalized Schedule and Create Experience Session Booking within this library of our actions. So once you find those, you can click Finish. So again, to get to that screen, we just went to This Topic's Actions, New and the globe, add from asset -- add from library -- Add from Asset Library and searched within this library for those actions we created and we're able to add it back to our topic. All right. So I'm going to go back to activate our agent. And I'm going to refresh this conversation preview. All right. So let's try again and try asking this agent about this full moon beach party. All right. So again, the agent is going to ask for my e-mail and member number, which is a great response. And I also see the actions now has 5 next to it instead of 0. So I know that this agent has the tools to respond to me appropriately. So I feel comfortable moving forward with this prompt. So I'm going to enter my e-mail address and member number. So we have this linked in your guide as well to make sure that you'll be able to talk to this agent. So if you go down to step 13, you'll find this e-mail for Sofia Rodriguez and our member number. So now that I shared my customer details with this agent, it's going to share with me the actual experience about what the full moon beach party is. So if you see on this left-hand side, these are the kind of behind the scenes of what action this agent is referencing. So it used the Get Customer Details actions to validate that I am a customer of Coral Cloud and then went into getting the experience details about this full moon beach party to share an appropriate response for what this experience is. All right. So I'm also going to try asking this agent more questions. I'm going to ask what experiences would this agent recommend. So knowing the actions that we created, this agent should then be able to generate a personalized schedule tailored to me and my interest based on my customer details. So pulled that generate personalized schedule action and created this whole schedule for today of different activities that I can do. So these all look awesome. I would love to be doing any of these today. So I'm going to ask to book one of these. I'm going to try the Beach Yoga Retreat. All right. So I'm going to ask to book the Beach Yoga Retreat and looks like there's some sessions available. The agent is going to ask how many guests I want. As you can see, it's using that Get Sessions action. I continue forward, it's going to use the create session booking. And we can really continue to interact with this agent and ask more questions, ask for details on water activities. We can also preview what this chat interaction is going to look like on Coral Cloud's website. So you can see from a more customer-focused perspective. But for the sake of time, we're actually going to stop here in this part of the workshop. So if you want to see what this looks like, those instructions are all provided in the guide. So after using Agentforce, we've given Coral Cloud a service agent that can help to handle that uptick in customer cases and answer those different questions that come through, especially about making reservations for these different experiences. So now I'm excited to hand it back over to Kiyoshi, who's going to analyze how this new agent is going to impact Coral Cloud.

Thanks, Shauna. Incredible job to you and to everyone following along. We just built our first agent. I know that's not Shauna's first agent, so I appreciate her leading us through that. But that was a lot of fun to see just how quickly an agent can be built with Agentforce. So once again, to recap, now that we've walked through building an agent together, let's fast forward, just as Shauna said, we made our reservation agent. It's been reviewed and activated. How is the agent performing? More importantly, how are we performing for our customers at Coral Cloud Resorts? Let's take a look. As a service lead at Coral Cloud, it's essential for me to help maintain smooth operations. Pulse's insights allow me to make informed data-driven decisions to enhance customer satisfaction and team performance. Through Pulse, I can easily monitor how many cases my team is handling at any moment. So just to orient everyone to the screen and what we're seeing here, this is Tableau Pulse. It's contextual, it's smart and it operates in a flow that's personalized to where I do work. So this could be in Slack, it could be Microsoft Teams, Salesforce, e-mail, you name it. And what's powerful about this is Pulse knows what metrics are important to me as a service leader. We're tracking things like average wait time, customer satisfaction, requests per agent per day and total cases. Now if I direct your attention to the upper left paragraph here, you'll see that nice sparkle indicating some AI that we're utilizing. It's actually providing a natural language summary that highlights for me as a service agent what is most important. So if visuals are something that I'm just getting more comfortable with, maybe I start with natural language or vice versa. The point is it's personalized for you. Now if I read the summary, it's saying that this month, customer satisfaction is up by 24% compared to last month. That's great news. Average wait time has dropped by 44%, perfect. Total cases have increased by 33%. So if I want a top line view, if I want a quick pulse on my metrics, I'm absolutely getting it here through Tablet Pulse. Overall, 4 of 4 metrics that I'm following have changed, 3 have changed favorably and one has changed unfavorably. Right. So let's take a closer look here and let's interact and drill in a bit further. It appears that cases have gone up in the last month. Okay. Higher volume of cases, not the best trend of the day. So let's drill in. If I click on breakdown, I see the Agentforce is already taking on a significant workload. The agent that Shauna just helped build and activated, it's already servicing those cases, and that's exactly what we want to see. So that's a huge support for the team. We can also track customer satisfaction, just as we looked at earlier in my digest, which from this visual, you can see that it's increased a full point, from 2.6 to 3.9. That's incredible. This is great progress. And beyond that, wait time and request per agent have decreased. All this is good. This reflects the positive impact of our Agentforce implementation. Now in the interest of time, I'm going to keep this next part very brief so that we can recap and have time for other exciting things and your questions. But I do want to scratch or pull back the curtain, I should say, a bit just so you can see how you create these metrics. So from here, I'll go to a New Metric Definition, connect to the data that I'd like to. Click on Connect. We'll give this a name here. We'll select a measure. So in this case, I'm going to be tracking talk time. That's something that we haven't done previously, can select an aggregation, a time dimension. So the way the Pulse works natively is working the time series. So let's see how this looks over time. And just like that, we're starting to create that Pulse metric. Now there's more that we can do here, giving end users the ability to slice and dice. So maybe I want the ability to add a filter here as this user is interacting with Pulse, things like that. Something that I'm super excited about is goals. So you can add in goals here. This is an optional step. But if you're approaching a target or you'd like to reach a target, this is a great way of measuring that success. And then finally, Insights. Now part of the magic of all of this is the metrics, but another part of it is the semantics, the information about the data. Well, in this case, when talk time goes up, that is unfavorable. So this helps us codify, if you will, our metrics in a natural language way so that we can recall this information in a way that we work with it in a way that's familiar for our business. Last one here that I'll just touch on since it has an asterisk. It's record identifiers. So if we're looking for outliers, that's something we can do here as well. Again, this is an optional step. But really, that's all that's necessary to be able to set up Pulse. So pretty impressive. Let me go ahead and stop sharing my screen and summarize essentially what we just saw, right? We accomplished with Coral Cloud Resorts, we used Tableau Agent, Agentforce and Tableau Pulse to give the Coral Cloud team the tools they need to optimize their businesses and improve the experience for their customers. I think that's pretty incredible. So with that, that wraps our journey with Coral Cloud Resorts today. Up next, I'll pass it back to Shauna, who will give us a glimpse for what's to come with Agentforce. Shauna?

Awesome. Thanks, Kiyoshi. So our workshop showed you this intro of how to build an agent, right? But if you spend a bit more time in Agentforce, you're going to be able to do so much more with it. And we're also adding these prebuilt analytics skills like Jeremy shared in the beginning of the session. So these are those Tableau analytics skills that help you with things like preparing and visualizing and analyzing data and also enabling that trusted Q&A. So we're going to take a look at what this Agentforce experience is going to look like with these new skills. So I'm going to go ahead and share my screen. All right. So let's imagine that I'm a brand manager for a company called DataFam Kicks and I'm going to try to hire street teams to get the word out about my company. So I have a ton of people that are applying to be a part of this team, and I want to make sure that I apply data and also my business rules to be able to manage these applications and really make sure that I'm hiring the right people for the team. So here, I have this application for Simone. So her application has gone into this data set in the spreadsheet of all of my job postings, all of the people applying and I really need to sift through all of these in order to find the best candidate for this role. So what would be really great is if I could take the power of Tableau with the data and trust and then also pair that with more agentic capabilities from Agentforce to make sure that I find the right people for this job. So I'm going to walk us through how we would do that and also create an agent that we can leverage using data. So I'm going to hit that plus New Agent button to start creating an agent. And I'm going to start from scratch with creating this agent in Agentforce Builder. So here, I can tell the agent what its role is just using natural language. I don't have to use any code. I can just tell this agent, hey, you are a recruiter, and you're going to review job requisitions, analyze profiles, explain fit and also make recommendations to our hiring managers. I'm going to click Next. And when we do that, it's going to go and find these topics for an agent that we can leverage. So you'll see it suggesting these Tableau-specific analytics skills like DataPro and also data concierge. So this makes a ton of sense because there's some work that we want this agent to do with analyzing and processing data and also that ability to answer questions from me and my team. So in data concierge, if I click on edit, I can tune exactly which of these actions I want this topic to be able to do for this agent. So I know I want for my agent, my colleagues to be able to talk with this agent in Slack. So thinking through these different actions that are coming up under data concierge, I think having use semantic data model is going to be coming into play and then also this generate visualizations, I want to keep that in as well. So I'm going to go ahead and click Next. And then we can see that it's taken my pretty minimal prompt of what I want this agent to do and created a really comprehensive set of instructions here. And I can edit these instructions if these don't seem right to me, but looking through what this agent created -- or Agentforce has created, this all looks really great to me, and I think this is going to work great for my agent. I'm going to go ahead and click Next. And then here, I can decide how I want to interact with this agent. So me and my team work in Slack. So I'm going to select Slack, but really any of these different channels would work just fine. All right. And then -- so what we've done so far is we've started creating our agent. We've added topics and actions for this agent. And then here, we have the real power of Tableau, right? So we can add Tableau data sources to this agent and we can also add things like Google Docs or more qualitative info like the company culture. We want DataFam Kicks to be building. So making sure that we have the right data checked here and then click Next. And we can review all of the topics and actions we have for this agent, the data sources we have selected. This all looks great. So we can go ahead and create our agent. All right. So now it's time to validate that this agent does what we want it to. So in this environment, we can ask a pretty simple question like how many active candidates do I have for this new role? So I'm going to ask the agent. So this seems like a really simple question, right? But when we say active candidates, that can really differ from one company to another. So one company might define an active candidate as anyone who's gone through a phone screen, but another might define it as someone who's gone through a phone screen and also has gone through a panel interview and perhaps is also someone that we've talked to in the past week. But what's most important here is that this agent has the same definition of an active candidate that me and my team and DataFam Kicks has. And I know that it does because I shared those business rules with this agent, and this is all trusted and backed by certified data in Tableau. So this looks great. It shared this great visualization with me. So I think this agent is ready to go. So I'm going to go ahead and activate this agent and see how we can actually use this in Slack. All right. So now we're in this Slack channel, and this is where me and my team work with the recruitment agent on getting these brand ambassadors for our team. So I can simply ask this agent in Slack to see who are the top 3 candidates for this role that we're hiring for. And when I ask, this agent responds with 3 candidates that it sees as the best match, and this is a data-driven recommendation. So I see this 92% match. This is a really high match, which I'm excited to see. That's awesome. We have such a strong candidate. But I also want to just understand what went into this such a high match score and this recommendation because we need to be able to trust and understand what went into this. So I'm going to ask this agent, why is Simone the best candidate for this role. I'm going to ask the agent to respond. And when I look in this response, the agent shares all this great info on Simone's candidacy. We've got some great qualitative info about who she is as a candidate. This is super helpful to see. But honestly, I'm still really curious about what went into this 92% score, and this isn't really sharing that. So I'm going to ask for more details. So I'm going to see if this agent can break down the score. All right. So here, we're going to get more quantitative info and a breakdown because the agent is going to query our data from Tableau and check our trusted data. So we get this breakdown. We also get this great visualization to see what went into this 92% score. So this is really just a glimpse of what it can look like when Tableau and Agentforce come together, and I'm excited for all of you to be able to experience it as well. So with that, I think I'm going to hand it back over to Kiyoshi, and we can open up the floor for some Q&A.

That's right. Thanks, Shauna. The future is bright indeed. On behalf of all of us, thank you so much for your interest and participation in today's workshop. We hope we included something for everyone, whether you're just getting started in Tableau, building your first agent or you've been with us for a long time, thank you for your partnership. We hope the demos and hands-on exercises were valuable for you. Now as Shauna said, we'd love to address any remaining questions you have. So please feel free to use the Q&A box at the bottom of your screen and myself and Jeremy and Shauna will do our best to provide clarity and ensure you leave with everything that you need. [indiscernible] some of those questions.

Yes. Boy oh boy, we've received a lot of questions over the past 1.5 hours, and it's impossible to get to them all. Maybe I'll kick things off if that's okay. I see sort of a meta question about how Tableau Agent specifically fits into the broader Salesforce AI vision. Remember, Tableau Agent was what you saw in sort of the first part of this overall webinar. It is that amazing experience that accelerated data viz efforts, calculation efforts and a whole lot more. I would say this. Right now, a lot of organizations are trying to, as we sort of say, DIY their AI, right? They are building models. They are training them, they are deploying them and then doing all that in support or in service of specific use cases. Well, first of all, that's really hard to do. Second of all, it's incredibly time-consuming. Things are evolving very, very fast that by the time you hit a milestone, there's already a new model out or some new innovation out. In other words, it's hard to even get started with AI when the ground keeps shifting. Agentforce, coming back to what I talked about at the outset of this webinar, it simplifies everything. It helps you kickstart your AI transformation with those prebuilt, those ready-to-use skills, many of which are designed specifically for data and analytics and Tableau to get even more specific. You can use those skills wherever you need them. We showed you how you can use those skills natively in Tableau Cloud, but also in the new Tableau, formerly known as Tableau Einstein to accelerate various tasks, you can also use those skills with custom agents to deliver smarter, more personalized experiences. We began to showcase a little bit of what that looks like today and will look like in the future towards the tail end of this webinar. Let me just close off, I guess, on this comment by saying this, this isn't just about like making things easier. Obviously, that is a big part of it, but it's so much bigger than that. It's about unlocking new possibilities. It's about accelerating your potential as data and analytics professionals.

All right. I'm seeing another one come in. I'll take this next one. I think it builds on what Jeremy just described. So the question is, how does Tableau and Salesforce's strategy towards AI and analytics differ from other major players in the market today? It's a great question, first of all. Marc Benioff has talked about this movement that we're in and this journey that we're all on together with the advent of LLMs and the acceleration of that technology. One of the ways that he describes it is in 3 steps, and I like 3 as well. It's the data, the model and the UI. So with Tableau combined with Salesforce, we have a lot of data. We have trusted data, and it's well supported by Data Cloud. So if you want to get the most out of both solutions together, Data Cloud is definitely something you should check out. Now all of that data, when you use it, it's important that you have trust. So within how we operate, we automatically have that built on top of the Einstein trust layer to give you confidence that you're working with your data in a way that's secure for your enterprise. Now I'm going to go up a level if you're checking with that 3-layer cake in the model. So just as Jeremy said, the space is evolving and changing quite quickly. So our strategy with that is to give you flexibility with the model. So within the Agent Builder that you saw in the Agentforce, you have the ability to specify which model you want to use, maybe it's for cost reasons, maybe it's for other preferences. But we want to give you that choice and flexibility while always maintaining that trust that we deliver on that base layer. And finally, the UI. That's the top layer of this cake. And the UI in Salesforce anyway could be Salesforce CRM, it could be Slack, it could be Tableau, it could be mobile, e-mail. Really, we're looking to serve you wherever you are, wherever you work so that you can make better and more trusted decisions. So Hopefully, that addresses the Tableau and Salesforce strategy towards AI and analytics. But as you just saw in the demonstration that was forward-looking for where we're going from Shauna, I'm super excited with what we have today and where we're going in partnership with you all. All right. See if there's some other questions here. All right. How can we ensure one version of the truth across our different teams as well as various technologies like Tableau and Salesforce? Great question. Ensuring that you have trusted data is always important. You just don't want to interact with an open LLM and be asking questions. Really, what we're describing here is a way to do it in a trusted enterprise way when you're working with your data. Now as far as one version of the truth, within Tableau, we have data management capabilities and certification processes. So I would say that addresses part of the process of that. In addition, on the Salesforce side and with Tableau, the new Tableau, we'll be able to utilize Data Cloud quite readily. And that will help us ensure we have one version of the truth as well as harmonize data with our Customer 360 information that we might want to take advantage of. So a couple of different things there. I'm describing more so the technologies, but obviously, people and process are going to be part of that as well. There was one -- so just taking a look at the clock, maybe I'll hit one last quick one, which is, how do I turn on Tableau Agent in Tableau Desktop? The answer is that you don't, you're going to be utilizing that on Tableau Cloud. So if you're looking for Tableau Agent, Tableau Desktop, you didn't see it. Thank you for loving Tableau Desktop. I love it too. Tableau Agent is available on Tableau Cloud through our Tableau Plus offering only. So if you're interested in that, I definitely have a conversation with your rep, and get you some more information on that. All right. I think that's about all we'll be able to address for today's webinar. So in wrapping up here, I just want to say thanks again for joining today's webinar. It was such a pleasure. It was a delight for me to be able to meet virtually with you all. Thank you, Jeremy. Thank you, Shauna. Thank you our marketing team. Thank you all for participating. Now before you go, we'd love to hear your feedback. Please tell us how we did with this presentation by scanning this QR code. All right. So safe travels to those that are off to your next adventure or to your next meeting. Thanks again for your interest and active participation today, and cheers.

Thanks all. Bye-bye.

Bye.