Keynote | Agentforce World Tour DC

It's not just what is possible, it's what you are going to make possible. It's a transformational moment. AgentForce.

AgentForce.

We're not just reimagining software, we're reimagining everything. It's called the Agentic Enterprise, powered by AgentForce. This is an agentic revolution. Big companies are using AgentForce. There has been a massive pivot as We are trying to become an agentic enterprise. Agentforce is fundamentally changing the nature of the digital world.

This is the internet again.

This is electricity.

It's that level of change.

The rate and growth of innovation, it's awesome.

In the last year, Salesforce crushed the numbers and the operating margins expanded.

We are witnessing the greatest transformation in business history.

With the digital workforce, the game has completely changed.

We got business value immediately.

Agentforce revolutionized our business.

It was an incredible quarter, and a lot of it has to do with the customer success that we're seeing around the world.

Companies are embracing this new way of working—

humans with agents driving customer success together, together, together, together.

We just really think about it as giving knowledge workers superpowers. That gives me time to get out with our partners and interact with them on a human level. Salesforce has helped us do work quicker where it matters, in the moment of truth.

And you multiply that by the thousands and thousands of patients people we treat, how could you not want to do that?

With Slack as the interface, it's not just chat, it's not just collaboration, it's the front door to the agentic enterprise.

Slackbot is the most powerful AI tool that we use.

If you're running business without Slack, you're doing it on hard mode. Salesforce, the revenue's just terrific.

This is the next revolution.

More than 50,000 people have come to San Francisco from Dreamforce.

This, my friends, is the AgentForce 360 platform.

We are building $10 billion businesses in service, sales, and data, all accelerated by AI. We have delivered 19 trillion tokens to our customers. There is no better metric for engagement. We have the vision that this new agentic enterprise is a system of context, a system of work, a system of agency, and a system of engagement. This is the powerful moment. The agentic revolution, there's no question It is here.

Like all good things at Salesforce, this is just the beginning.

We saw 185% increase in number of appointments. We have seen a massive efficiency where we have implemented AgentForce in the range of 25 to 30%.

We're at 87% automation of the claims process.

Nobody else does that.

I think you push the chips in and transform your business.

We've gone faster with AgentForce than we have gone with any other technology. This is the agentic enterprise.

Please welcome President and CEO, Mission Force and Government Cloud, Salesforce, Kendall Collins.

Thank you, everyone. It is amazing to be here. Welcome to Washington, D.C. World Tour. We are just delighted to be here in the nation's capital with all of you. You can feel the energy. It's an extraordinary time in the world, an extraordinary time in the US, an extraordinary time at Salesforce. You know, we are just seeing the world change so quickly. I've been at Salesforce 23 years. I know I started when I was 11 years old, but I actually last did a keynote in this room 20 years ago. Somebody just reminded me in the backstage. And I'll tell you, there's just never been a more exciting time. There's never been a time when we have felt more innovation, more fun, and more opportunity. And today we're going to transform you. We're going to tell you about what's happening with the Agentic Enterprise. But before we do that, first, we have to thank you. Everything for us really starts with gratitude. We're grateful for the opportunity. Opportunity to serve. We're grateful that so many of you are here, our Trailblazers, our Golden Hoodie, our Golden Dome Hoodie, our online folks who are watching us and tuning in, our nonprofits, our entire community. It's just a huge part of who we are is expressing gratitude. So thank you. Thank you for being part of our community and giving us energy every day to do what we do and to serve you. It's a real privilege. Well, for many of you know, we are deeply rooted in our core values. We've had these values since the start of the company. Trust, customer success, innovation, equality, sustainability. We believe that our values create value. They're not just words. This is part of our code. This is ingrained into our culture, into our operating model. This is just a huge part of who we are. And, you know, I left the company for about 7, 8 years and I came back 3 years ago. A big part of why I came back were these values, because the values give us energy. They give us that passion. They give us the commitment to know that we have the best team working on the most important problems. And, you know, these values have also created value for our shareholders and We're very grateful for them. Can't believe it, but we're on trajectory to $63 billion in revenue. I never could have believed that when I started the company. We just had a few hundred people and, you know, maybe $100 million of revenue. And it's incredible. We've done it while being one of the most innovative companies, one of the 100 companies in the world that cares, and one of the world's most ethical companies. And all of this is possible, again, going back to those values, because of innovation. You know, innovation is one of our core values, but it's also fun. You know, we're looking at what's new in the world, and you think about everything that has happened over that 27 years, it's incredible. We now have tens of thousands of developers around the world innovating, writing code, working on the hardest problems. We've also done amazing innovation through acquisitions, whether it's Recently, Informatica. Welcome to the family, wherever you all are. I think you're here somewhere. Slack, Tableau, MuleSoft. It's been an incredible family and 27 years of innovation. You know, and we haven't just done that innovation and created a new technology model and a new business model, but we also did something else really important, which is We created a new philanthropy model. And this was such an important idea at the beginning, to take 1% of our time, 1% of our equity, 1% of our product, and to give that away and to make that the center of energy for the community. We've given away now nearly a billion dollars, which is incredible. $55 million here in the Washington, D.C. area. We serve nearly 65,000 nonprofits. You know, and volunteering is still such an important part of how me and my team get energy every day. Thank you to all the nonprofits here. There's so many great ones. AARP is here. Blue Star Families is here. Hiring Our Heroes is here. You know, there's too many to call out. And it's just incredible that you're all here. You know, but maybe nowhere are our values and our mission more intertwined than here in D.C. You know, we think about Washington, D.C. and this commitment to serving those who serve. We've been doing this for 27 years with nonprofits, but 20 years with the federal government. The U.S. government is our single largest customer, our single biggest, most important partner, you know, and they drive us in so many important ways. We've been supporting their mission, these public sector organizations, and so many of them are here, you know, public servants, many of you in the room, I just want to say thank you. You have some of the most complex missions. Thank you. The complexity of those missions is no joke. It's national security, it's public health, it's disaster response, it's making sure that veterans get their benefits on time. You know, these are complex missions and really important. We're so happy to have so many of these customers here today. Department of Agriculture, Department of Veterans Affairs, Josh Geiger, thank you for all the work you're doing. The Air Force, the U.S. Army, Andy Fell, good to see you, pleasure. Department of Transportation, the IRS, CMS, Centers for Medicare and Medicaid Services, you're going to hear from our friend Patrick Newbold, incredible CIO in the federal government, and over 70 federal agencies. And so many see other great friends here. Former Secretary of the Navy, thank you. Thank you, sir, for your service. Great to see you here. Thomas Corlett, EPA, a delight. And our friend Robert, good to see you, of the White House. Thank you for being here, sir, and for all your service. I just want to thank you all again because without you, this just really wouldn't be possible. You know, you give us the energy, It is a cause for many of us and a really important double bottom line. And, you know, I just want to take you through sort of what happened in my life because I've only been working with the government deeply and closely for the past year. And it really started for me a year ago. A year ago, Mark and I hosted a dinner in Honolulu, and we're in this dinner with so many important diplomats, and business leaders and customer leaders and military leaders. And the military leaders asked us in that meeting, they said, we would like you to do more. Can you help us modernize? You know, our warfighters deserve the best. And we just want to bring the best of the private sector, we want to bring to the public sector. 80% of what we do is not exotic. We want to leverage the best of what Amazon does, the best of what Walmart does, the best of what Delta Airlines does. We want to leverage that for our troops and for our national security. And we said, absolutely, it would be an honor to do that. And Mark turned to me and said, Kendall, this is your full-time job now. And we launched Mission Force about 8 months ago. It's our commitment to those who serve and to national security. It is our product portfolio and bringing the best of innovation and what we do to all of you who serve. And it is also specific products. Okay? To scale continuous mission readiness from insight to decision to course of action, whether it's personnel, logistics, analytics, operations, you know, we are here to support you. And the amazing thing is we have over, let's see, 6,000 veterans at Salesforce. Dean Robison, our executive sponsor for Vetforce, thank you for being here, former Marine. I think some of our Vetforce folks here, Tom House and crew, but to have the firsthand knowledge of what's required in these complex missions is great. And being able to leverage that with our team is critical. Well, we've been recognized, and we're humbled by this, for our commitment, but also for our innovation in public sector. And in many ways, it's almost not fair because government is really an industry of industries. Almost every industry plays some role in the government. So we've been building industry products for the past 10 years. We've got financial services and health cloud and automotive and industrial. We take the best of the best. We're able to leverage that for public sector in addition to functionality that we build for permitting and personnel, grants management, case management, all of the critical functions inside of the government, even tax and revenue management. And we're honored to be awarded, you know, this, this recognition as a leader by Forrester. For uniting program data, casework, and citizen engagement on a single platform. So where do we go from here? I think I'm going to go this way. You know, this is an incredible opportunity to come back to our vision since 1999, which is to connect with our customers in a whole new way. You know, I think that umbrella's gotten a little bit bigger. There's room under that umbrella for citizens, And taxpayers and veterans and all of you service members. And it's about connecting with you and serving with you and bringing that vision forward with the latest technology. And when we think about what's ahead, we've seen these tectonic shifts in technology. You know, obviously cloud computing, We started there. That is the principle that the company was founded on. I was very fortunate to be the general manager for mobile, the general manager for social. I'm still a bit partial to Chatter, but I love Slack. Okay? Predictive AI, we've been doing for a decade, processing trillions of predictions in AI. But now, agentic AI is here. It's undeniable. And agents are changing everything that we do. We feel that. We feel that as individuals. Do you feel supercharged? I feel supercharged when I'm using AI. If I'm writing code, I'm researching things. And let me tell you what's crazy. My daughter, she's 10 years old, and she used AI to create an online flower shop. 10 years old. So it's unlocking this incredible ability for you as an individual. It's supercharging you. Your creativity, your entrepreneurialism. She's like, Dad, I'm selling flowers. Dad, I'm going to start selling lemons. Dad, I'm going to start selling your clothes. I'm like, no, wait, stop. Stop selling everything in the house. But your creativity is incredible. And then what's interesting is we think about AI in the enterprise, and we know every organization is going to become an agentic enterprise. You know, we're starting to use AI now with our teams. We're driving better customer experiences, higher revenue, greater productivity. Humans there are coming together with the business logic that we need inside of a company. But then the question is, where is the government? Well, it's raising expectations. You know, the government must keep up. 75% of people expect government services to match the private sector. I don't know what happened to the other 25%. Did they give up? I'm not sure what happened. They haven't given up, have they? Dave Ray, our head president of public sector, amazing to see you here serving all these customers. They're not giving up and we're not giving up. Citizens are expecting 24/7 availability, instant expertise, personalized answers, and they need all of you. They need all of us in this room to come together and bring our energy and our ideas and innovation to serve them because we cannot open up a skills gap. We cannot open up a trust gap. We cannot have a chasm between what we experience as citizens from versus what we experience as individuals. We have to unlock this superpower for the government and for nonprofits and for education. It's absolutely critical to the country and to all of us. But we know from doing this the past 2 or 3 years that AI models alone can't run your organization. Who's been told here, oh, all you need is a model, just roll out a model, it'll be great, it'll solve all your problems? It's just not the case. You know, models can't act on their own. Sometimes they get things wrong. They can't guarantee outcomes. They lack transparency. Transparency. Sometimes models are confidently wrong. Does anybody here— do you know anybody? You have a friend, maybe a family member, don't say it out loud, and they're confidently wrong, sometimes about politics, sometimes about a recipe at Thanksgiving. There's nothing worse than having somebody who's confidently wrong and doesn't know your business, but they're really in your business. You know, that makes it even worse. And this is why 95% of AI pilots are failing before they make it to production, because the models alone can't do it. And we realize this through our experience with now thousands of customers deploying AgentForce. The agentic enterprise needs more than models. It needs 4 systems. It needs a system of engagement, a system of agency, a system of work, a system of context. These systems have to work in concert So that's why 2 to 3 years ago, we started rebuilding everything that we'd done at Salesforce, all of our work, all of our innovation for decades. How do we bring this into a unified architecture? And it starts with a system of context and data. And people talk about AI, but I don't think we're talking enough about the data. The data is the context. Being able to bring all of that together with Data 360 is critical. It's your memory. It's your institutional knowledge. It's the recency of what just happened. There's nothing worse than when AI or a call center or any service experience doesn't know what just happened to you and can't bring that context forward. It's a system of work built on top of that with a Customer 360 and all of your apps that you've been deploying across sales, service, field service, now ITSM. And so many different solutions across the application stack. It's a system of agency, a way to build and deploy agents, and of course, a system of engagement with Slack, and we're going to talk more about that. But I want to call out one other area, which is the models at the bottom. The models are important, but a key part of our stack is to bring your own model. We want people to be able to use the biggest frontier models, But we also want you to be able to use small models and special-purpose models. I believe, you know, that a lot of these big models are becoming commoditized in many ways. They're getting closer to the same capability. But when you think about a small model and you think about the data that you have in your organization that is unique, I believe many people are going to be training models specifically for narrow purposes. That data asymmetry is your asymmetric advantage. And I think you're going to see more and more of this. So, your ability to bring any model is absolutely crucial to making this all work. And we know that these AI models need trusted context. We talked about that a little bit, but one of the foundational principles is zero copy, right? You have invested a massive amount of energy to get your data into the right lakehouse, whether it's Databricks, Snowflake, Google BigQuery, Amazon, you name it, but you don't want to move it. You want to be able to leverage that, leverage your past investments, But you need that context. You need to bring it together to make AI and agents work. The other piece, which has been phenomenal for us, is we've seen Slack and all of these conversations that we're having, being able to bring that and that context to bear inside of the data layer. And now with Informatica, we're seeing the ability to bring quality and clarity in new ways, new products that are also here for the government, which is incredible. So, we know AI models don't know how to run your organization, but the apps do. You spent 27 years, some of you, I see so many old friends here, it's incredible. And you've been with us pouring your heart and soul, your energy, but also your logic, your know-how, the business rules, the workflow rules. You've been codifying it into the apps. You've been building journeys for your citizens and for your processes. Well, that doesn't go away. It's critical. It's critical that that is part of the stack to make an agent successful. And of course, delivering trusted results every time is all about getting predictable outcomes, right? How do you get to a predictable outcome? Well, let's start with the concept that there is a right and wrong answer for some things, even though AI as a model may not believe that there is a right and wrong. So we started working on something and we call it determinism. And its agent force script. But this concept that you can write a script, write a rule, write a policy, and unequivocally know that you're going to get a trusted answer is critical. So that was a key piece. The other key learning that I really had over the past year is if you want trusted AI, you need observability. Observability means you have the ability to open up the agent, look inside, under the hood, How is it making a decision? And we're going to show you that today, because then you trust it, then you can tune it, then you can fix it. So this is incredibly important. And of course, Slack is where the agentic enterprise works. Slack, we're living in these channels. There's so many messages, sometimes a few too many messages, but that's okay. We have Slackbot, and Slackbot is an incredible ability to have your own personal agent. To get signal through the noise, to ask it anything. It's not just search, it's summarization, it's creating new content. And when I'm in Slack, I'm doing approvals. I approve expense reports in Slack, I approve POs in Slack, I'm doing performance reviews in Slack. But we've done something else, which is we've introduced agents into these channels. You know, you can take a PDF, you can train an agent on a specific policy, and put it in a channel, and its only job is compliance. We've done that internally. We have an agent for moderation. We have one channel, some of us know it gets a little spicy, you know, that channel and some of the employee conversations, and that's okay. But we have an agent for moderation. We have an agent for project management to keep us on time and on track and keep nudging us. This has been so amazing internally, and we've really had to transform our entire company from the ground up. It starts with us being Customer Zero and Salesforce being an agentic enterprise. You know, it's not just using Slack. It's 2 million inquiries that we have now resolved through customer support and agents. And many of you experience it as customers. You come to help.salesforce.com, you ask an agent a question, and it's there 24/7. You can always get to a human, of course, but this has saved us hundreds of millions of dollars, money that we can put back into the product, back into innovation, back into so many things that help serve our customers. So we see it elevating our employees, elevating our customers, and allowing us to reinvest in key parts of the business for our customers. And this is just a phenomenal set of learnings that we can now bake into the product. But it's not just us, it's all of you. Thousands of customers are driving value with AgentForce. The U.S. Postal Service, a customer package is missing. Where was it last scanned? Incredible agentic process. National Institutes of Health, what's still needed from this participant for the research study, driving public health outcomes and accelerating research, accelerating care. Blue Star Families providing family support to our service members, to our veterans, you know, figuring out programs near your base. 24/7 support is the new normal and is so crucial. We are delighted about all of the learnings, everything that you're pushing on. This is the moment. This is the opportunity to take advantage of agents and AI. And we're seeing our customers. We're going to hear from AARP, PenFed, Dell. You're going to— we're going to take you behind the scenes, the Centers for Medicare and Medicaid Services, and have a great conversation about it. So please welcome to the stage. Let me bring up Avantika, who's going to take you through AARP.

Awesome. Thank you, Kendall. Good morning, DC. I'm thrilled to be back here for the second year in a row since last year's world tour. Now, as the largest nonprofit dedicated to serving Americans over the age of 50, AARP supports over 38 million Americans through some of life's most important and oftentimes most vulnerable moments. So when a caregiver calls AARP, it's usually tied to a high-stakes situation related to health, safety, or an urgent matter for a loved one. So when they call AARP, it's critical that they receive clear and immediate guidance, because when they don't, that means they're left with uncertainty and time and attention is pulled away from those who need it most. Well, thanks to Agentforce Voice, AARP is able to provide intelligent support 24/7 to their members through channels that they trust most. In fact, why don't we go ahead and check out a Visionary Experience of what that feels like for an AARP member. It's taking a moment to connect to ARP. Should we give that another try?

All right.

All right. This one you know is real. Let's give them one more try. Yes, maybe a different phone will do the trick.

Hi, I'm the AARP customer agent. Thanks for calling AARP and thanks for being a member. How can I help you today?

Hi there, I recently became a caregiver for my mother and she's actually moving in next week. And I need to make my home safer for her, but I have a lot of steps near the entrance of my home. What do you recommend I do?

Margaret, that's wonderful that you're taking steps to make your home safer. I'm going to send our Home Fit Guide to your email on file. It has a lot of helpful information about improving home safety.

All right. Thank you so much. All right. What do we think? It was good. It was fine. Agent Forest was able to recommend the right PDF for Margaret to read in order to get her question answered. I don't know about you all, but if my mom was moving in next week and I have to prep my house for her, I don't know if I have all that time to read through a massive PDF. Why couldn't Agent Forest just give me the specific answer I needed from that guide? There's a reason why. Now, if we look at the HomeFit guide that AgentForce recommended, it's a 36-page PDF filled with diagrams, visuals, tables. And when this gets sent to the agent and the underlying large language model, it gets flattened, meaning it gets converted to raw text like what you see here, which is much harder for a large language model to make sense of. And this is why Agent Force was unable to provide that specific answer that Margaret needed. And this is the problem that Intelligent Context has been designed to solve. So what we're going to do now is go ahead and upload that guide to Intelligent Context. Now, by a show of hands, who here has to deal with searching through PDFs on a daily basis? I'm sure all of us. I mean, look around the room. Here in this city, in DC, we deal with a Potomac of PDFs flowing through our systems every day, and we know just how critical every single page of that PDF is important for context. So now that Intelligent Context has processed this guide, you'll notice that it's generated chunks. Chunks are essentially meaningful slices of this document. And this is how the structure and relationships between all that data is preserved. So why don't we go ahead and ask AgentForce that same question that Margaret asked earlier and see if it comes back with a response.

There we go.

Now AgentForce was able to extract the right answer for Margaret from the guide. So it recommends that she installs a ramp and adds handrails This is great. But there's one thing that we need to catch here. If we look closely at those generated chunks, it's actually exposing some PII. It's exposing the names and contact information of internal employees. And this is happening because Intelligent Context is processing multiple documents, including internal documents. So we need to make sure that that data is masked. And this is where the picture. So with Data 360 Governance, you can actually define custom policies for your agent. And Agent First makes it even easier where you describe it in natural language and it automatically gets added. And then once we activate that, you'll see that it's redacting that sensitive data. So, so far what did we do? We went ahead and uploaded that home fit guide. We made sense of the data through intelligent context and we masked sensitive information. We need to go build our agent, and this is where the new agent builder comes into the picture. So here's our new agent-first builder. What I'm going to do here is I can actually start from pre-built templates. So organizations like AARP can get started quickly with topics like volunteer management and volunteer participation and go live fast. Now, I want us to double-click into this topic right here around providing Guided advice to caregivers. You'll notice that we actually have explicit instructions to ensure the agent never answers any questions related to providing medical or occupational therapy advice. And the way we do this is through this conditional expression. And this is what ensures a level of determinism. We're injecting this deterministic logic to ensure the agent never deviates from non-negotiable instructions. And this experience is powered by agent-first script, which is going to be critical for industries, regulated industries, and organizations like ARP to ensure that consistent and compliant experience. Now, as a final step, let's add that context to our agent. So we're going to go ahead and add the HomeFit guide that we just processed. Through intelligent context. We're going to activate our agent. Why don't we go ahead and give that call one more try and see what we come back with.

Hi, I'm the ARPE member agent. Thanks for calling and thanks for being an AIRPE member. How can I help you today?

Hi there, I recently became a caregiver for my mother. She's moving in next week and I need to make sure my home is safer for her, but I have this huge staircase in front of my home. What do you recommend I do?

ARPE is home fit. Suggests a few helpful entryway updates. Add sturdy handrails on both sides of steps, reduce high thresholds, and improve lighting near the door. You might also use non-slip surfaces, widen the doorway, or add a covered area for protection. Even small changes like these can make entry safer and more comfortable.

Awesome, super clear, super specific. Thank you so much. All right. How was that? Much better. And this is the new experience that AgentForce and intelligent context is really powering. And now you'll notice that Margaret no longer has to spend that time searching through that massive guide. She can channel that energy towards really caring for her mother. And you'll notice along the way, the agent never compromised safety, accuracy, or compliance. With these grounded agents and that trusted system of context, this is AARP's Agentic Enterprise in action. With that, I'm incredibly excited to introduce my friend Corey Haines to share PenFed's narrative.

Awesome.

Great job, Avantika. That was awesome. Well, it's great to be here in the DMV. For those that aren't familiar with that acronym, that's DC, Maryland, Virginia, not the place where you get your license. So it's great to be here in this great season with the cherry blossoms. And I'd love to talk about PenFed, one of DC's very own, the DMV's very own. But before I do, I'd like to give a shout out to my demo drivers over there, Gus and Jackie. Look at them, they're low-key, aura farming and Looks maxing over there. That's what the kids say, you know, but they're looking great over there. So let's talk about PenFed. You know, PenFed originally, they were offering military-grade services, financial services to the men and women in uniform, which is great. And in 2019, they wanted to expand those services nationally. And to do that, they thought AgentForce would be great because they have to keep the same headcount and they were really going to make HR and IT more productive. But they realized that they had to do more complex services, so they need more specialized agents that could offer orchestration. And with that, they needed more advanced data models, things like agent-to-agent protocols, MCP, industry workflows, and data models, which were great. But they also now understand that they had to really automate some of those things so that they could focus on the how, so that their team members can focus on The who, which was growing their member services. So they realized that we wanted to show you today, we want to show you a demo that shows you how these agents have full orchestration and this orchestration is going to go in the flow of work. So first, I'm going to show you one of the great IT agents, Penny. It's a pretty cool name, right? Penny. PenFed uses this IT agent Penny and she does about 75% of the IT tasks, mundane IT tasks like password resets, onboarding, things like that. So the IT team can focus on really mission-critical IT tasks and things that really need human expertise. So now the IT team can focus on those things. But the great thing about it is that as great as Penny is, sometimes the employees still have to go find penny, they still have to go to the IT system and they have to really leave their flow of work. So it's not as productive as you would think. So what if they could just stay within their flow of work? What if they could do— find the IT task and stay within their flow of work? So I'd like to show you something a little visionary, show you something that shows us in the flow of work. So I'm going to take a moment and I'm going to switch gears. I'm going to be a member service agent. So I'm going to switch gears for a moment. For a moment, imagine me as a member service representative at PenFed.

Okay?

Put your thinking caps on. I'm a member service agent. So I'm this member service agent, and I'm taking a call. And I'm taking this call, and I'm tracking all the things in the call, I'm taking notes, I'm summarizing the call, doing all the things. And as I'm doing all the things within Slack, I'm even updating the call record for Salesforce within Slack. I'm not even leaving Slack. Usually, I would go to Salesforce to update that, but I'm going to do it right there in Slack. And as I'm doing that, one of the things that I had a conversation with this customer about was that they wanted to upgrade it to a premium savings account, which is great. So I'm going to update that. And as I'm updating that premium savings account, I'm going to put that information in. One of the routines is that I need to have a member service representative help me. And this member service representative in PenFed is an agent. It's called Wingman. Cool name, huh? I guess PenFed has these great names. So Wingman as a member service representative, as an agent, is going to help me through this process. And as I'm going, wait, Wingman tells me I don't have sufficient permissions.

Huh?

And the reason I don't have sufficient permissions because I, as a representative, need something more. So just like a person, the agent is telling me I don't have permissions to do this task. So what the agent is going to do is going to route me to the right approver to make sure that I have the right permissions. It's going to route me back to Penny. Now, take a moment here. This is an agent working with another agent to route me to get the right permissions. Agent to agent doing a routine. Also note the trust issue. They're following procedures, the PenFed guardrail in place. They're making sure that the policies that they're following, they're not going to break any KYC policies, they're not going to break any policies that PenFed has. They're making sure that I have the right permissions. Everybody in here knows who works for government or regulated industries, you don't break any permissions or policies. If you don't have clearance, you don't get in the building, right? So we see agent to agents working together, we see permissions happening, we see trust happening all within the flow of work. Pretty amazing. So as this routine is taking place, agents are working with agents, let that happen. Now, our IT team can focus on mission-critical task. So let our IT team, looking at our— they're going to focus on what they need to focus on. So now our IT team can focus on— they're looking at their IT service screen, and they see some— a lot of tasks here. And look at all these things that are going on, and they seem to be a lot, but they can't go through all of this, right? It would just take too much manually going through each and every ticket. So they're just going to ask AgentForce. They're going to ask it right there in Slack, what's going on? What do I need to focus on? And they're going to ask it to say, look, there's a pattern that we see. What's the pattern? There's a VPN connection, there's an update that went in, and there's something that went out. What do I need to do? They say, hey, if you just roll back this update, we'll get the fix in, and we can notify all the employees. Well, let's do that. This is how the MCP connections come in, right, that they're able to do all that within Slack.

This is pretty cool.

They're taking all that right there. Now, this is happening in Slack, Now, we know everybody is not a Slack customer, and we understand some of you have heterogeneous environments. Can happen in Microsoft Teams as well. We got your back. We understand that. So think about this. We just saw Penny, the IT agent, take care of an IT task. We saw in the flow of work Slack take care of all the agents working. We saw IT services doing and resolving some complex tasks. That's a lot of things happening. Agents doing a lot of work. How do we trust it all? How do we know all of this is working? It's almost like, you know, when my kid says he took out the trash. But did he? I don't know. Do you really trust that? And you heard Kendall talk about earlier, you know, we got a big trust chasm going on, and I know here in D.C., trust is a big issue. You guys have a thing that you guys say, you've been saying it for a long time, trust but verify. This is where agent force observability comes in. You want to make sure that everything you're doing and these agents are doing is really happening, right? This is where if we look back at Wingman, we can go back and see step by step what was happening. We can even look back and see what we need to improve on. We could see, when Wingman did it, how Wingman did it, more importantly, why. Now, I work in financial services, and if you get audited, you need to definitely tell the why. You may get audited tomorrow, you may get audited a month from now, 7 years from now, but if you have that why answer, it's going to be critical, and you need to be able to produce that information. So this is very important. So having that information is critical. So now that you have all that information, this is why we think you have now this trust factor. I want to leave you with this information, and this is so important to me. I talk with financial services leaders every day, and they tell me, look, this is all great technology, but it's not just about deploying agents. It's having agents and humans work together to solve complex problems in regulated industries in secure environments at scale. PenFed is rewriting the way they get work done. That's what's most important here, and that's what truly separates them to be a true agentic enterprise. I hope you've been inspired today like I have with PenFed. And with that, I'd love to tell another story but I can't tell it. I'm going to hand it off to Maryann. So, Maryann, tell us a story about Dell.

Thank you. Thank you, Corey. Thank you, Corey. Dell has made a huge bet on AgentForce to transform how they are working. Now, before AI, Dell's critical operations, they required armies of people sending emails, chasing approvals, collecting information. And this was slow, it was expensive, and it was really error-prone. Sound familiar? Well, today with AI, Dell now has 19,000 users that are using AI to transform the way that they work. Automated workflows, digitizing processes in minutes, not months, orchestrating work across humans and agents, and keeping the business in control. That's real transformation.

At real scale.

But the Dell problem, it's not just a private sector problem. It is a people problem. And I want you to think about your best people. Are some of them maybe stuck in forms doing a lot of manual work? Is that why they came to D.C. to help all of the millions of Americans? No. That is something that we can go automate with AgentForce. This is the work that takes away from the mission, and we want to give you the time back so you can focus on the mission. So big thanks to Lynn and Jackie here. They're going to help with me with a demo, and we are going to talk about Dell's supplier onboarding process. Now, the great news is to get started, you do not need a perfect document. In fact, I have here a hand-drawn document that shows Dell's supplier onboarding process. I have some information that I'm going to be gathering, a certification review, There's some business logic, and then I have some ideas on how agents are going to help. And Jackie's going to take a picture of this and upload it for me. Thank you. But, you know, a document, a photo, a napkin sketch, that's all you really need. And as we get this started, I'm going to take you on a quick thought experiment. It's a live demo, so going to kill a little bit of time. And I want you to think about your most complex processes. I bet you have a bunch of branching logic. There's a lot of forms, there's a bunch of information you need to go collect, and hopefully there's not some business logic that walks out the door when that person retires. Well, we can automate all of that. And I see your face and I know what you're thinking. Hey, lady, I know you got this AI and you think it's pretty cool, but I have security and compliance concerns, and I'm not just going to trust AI to handle a process and like, hope and pray that it works. Well, we got you covered. Trust is our number one value. And so that's why everything in Agentforce has fully auditable AI. You can decide where you want to keep humans in the loop at every process or no step in the process. And we have air gap support for all of your most sensitive workflows. I got another good piece of news for you. This also integrates with your legacy ERPs and maybe some of those systems running COBOL. I hear the Social Security Administration has some of those, but I'm sure some of you in this room do as well. Okay, let's go back to the demo. It's done a lot of hard work. It took my process, broke it down into steps, took each step, assigned it tasks. For every task, it then looked, should this go to a human, Or should this go to a specialized back office agent? And now all of that work, it's all captured right here. Every step, every stage, every process fully automated and also keeping your business logic in check and flagging any exceptions. Because in the agentic enterprise, Agentforce is orchestrating your work. And I want you to look at where Jackie's, mouse is right there. In this case, this workflow was directly assigned to an agent. We have agents that are extracting data from documents, doing tabular calculations, having complex routing logic, and we've intelligently inserted where all of those agents are orchestrating the work. Uh, this is amazing. Um, so let's switch to the future And look at what this might look like inside of Slack. I wanna make sure that my teams have all of the information. And so here we have one connected point of view. Everything that you need is all in one place, and your team has everything that needs. And so when we click in, we can see all of those steps, all of those processing. My team gets everything that they need right in one view, right when they need it, so they can see everything that is in motion. Now, this process automation looks great, but I want to leak you a little secret. I actually think the public sector is better set up for process automation than almost anyone. I see a lot of blank stares and you're thinking, why? Well, it is because you have so much documentation. You have process documents, you have compliance manuals, you know what good looks like, and thankfully you wrote it All down. And so Dell has the same discipline that you do, and that's what enables them to have really great measurement of what they're doing. Here, the supplier onboarding process looks like it's taking about 20 days, and Dell can go in and ask the Tableau Analytics agent, what can I do to help improve this? And what I love is that it goes right down into the data. It looks through that process and it's thinking, How do I improve it? Well, here you can see the bottleneck is that the financial document verification step should only take about 4 days, but it's taking 9. Now Dell knows exactly where to go, iterate, deploy an agent, and automate and accelerate this process. You have the exact same tools. You can do the same thing. Measure, automate, iterate, improve. Everything that you need is right there. I just want to end with a quick message. DC is with some of the most talented, dedicated people in the world. You took your job not because you loved forms. You took it because you had a mission. You wanted to serve your community. You wanted to serve your neighbors, and you wanted to serve your country. Let AgentForce do the automation. Agents are great at manual work. That way you can go focus on the mission. We're deeply honored to be a part of your journey, and thank you for listening today. Back to you, Kendall.

I don't even know where Mary Ann is. Amazing job. Amazing job, Avantika and Corey, and for sharing those incredible demos. Everything we're showing you is obviously real code. We take a lot of pride in the innovation. And I just want to say, it's so energizing to see that, you know, the opportunity for process automation across public sector is incredible. My dad was in the Army. He used to say the Army has got a manual for everything, literally everything. You know, there is a PDF, a Potomac of PDFs out there. You know, as you heard from Avantika, being able to bring those to bear and do the automation is so critical. And all of that is so that you can have humans still in the loop, but not just in the loop. I'm going to steal a line from our next speaker, humans in the lead, you know, and humans in the lead designing processes, ensuring trust and verify that the process is right, ensuring that there is a right and wrong answer. And that citizen, that patient, that veteran, that service member is getting exactly what they're supposed to get. It's so critical. And we're just delighted to hear from our next speaker. Please give a warm welcome to Patrick Newbold, the CIO, Centers for Medicare and Medicaid Services. Please join me up here in the cloud. How you doing? Welcome to world tour. It's amazing to have you here.

What a sight.

So, you've had an incredible career. You know, Patrick has been the CIO not just at CMS, but Social Security, NASA, you've been in the Army, you've had an incredible career. First off, just thank you for all your service as a public servant. It's incredible. I think for everybody here, they'd love to understand, you know, you've been on a journey with AI. Maybe tell everybody here, like, what have you learned? What has been, like, your journey so far at CMS? And what's been the hard part? Like, what's been the hard learning that everybody here can benefit from?

Yeah, I'll just say one is, what you alluded to. I think it's very important that the humans are in the lead, setting the context, setting the objectives, setting expectations, not just in the loop. And I think that's something that we have. The second thing I will say is that when you're adopting AI, it's bigger than just the product of tool. You have to bring what I call the total package. So that's the enablement, That's the training, the education, so that the workforce can use it to drive some outcomes that we talked about today.

Yeah, I took that away from our conversation, the total package.

Total package.

Because in and of itself, you know, model is not enough. You need a unified stack, but you need this total package with enablement and training and kind of all of that support infrastructure. You know, you don't have an easy job. You got what, half of Americans?

Yeah, yeah. 160 million.

160 million Americans. No big deal.

No, no.

Small trillion. $1.8 trillion.

You know, if we were— we'll be the 13th largest economy if we would measure that.

13th largest economy in the world and probably need some automation. Maybe tell everybody here, when you think about AI and where it's all going, like, what's next? Where do you take it from here? Because you've done so much. To automate already, but what's next?

You know, I think the, the one of the biggest areas that we really want to focus on is fraud, waste, and abuse. We want— we, we have a goal to crush fraud. And when you're talking about paying out $1.8 trillion a year, um, even 1% of that is $1 billion if you can avoid it in fraud.

Billions.

I love a billion-dollar ROI.

Well, I think the American taxpayers will love it too, because that's more That's more funding, that's more resources that can go towards care. And that's our goal.

Yeah. Well, as taxpayers, we're counting on you for the billion dollars.

No pressure.

You know, for those of us who have family members rely on your critical service every day, like, we're counting on you. And, you know, we can't wait to see what you do with AI. I know it's been an amazing journey, and I'm just so grateful for your leadership in it. Thank you.

Thank you.

Thank you so much, Patrick. And thanks, everybody. I think we have a— I think I got a whole row of CIOs out here, which I'm excited to spend some more time with as well. So I know so many of you are doing critical work. Well, let me say this. The agentic era for government, it is now. You know, we're excited, we're having fun, we're doing innovation, we're bringing it to all of you. We've got these products available today for FedRAMP High. You know, the demo you saw from MaryAnn, we just are so excited to bring Mission Force Process Automation to bear on the public sector. You know, we know many of you have secure environments, air-gapped environments. You know, supply chain is one of these things in the private sector that needs to be air-gapped because of security reasons. So we're delighted to extend that into the government and to all of you. And AgentForce IT Service, you know, having a change management database, having incredible ITSM, this is something that we know all of you deserve. And want and have been asking us for for years. So we're delighted to be bringing that this year to all of you, IL5, and marketing and public sector. And of course, Data360 and Tableau Next. You know, it's an incredible time. And I want to wrap up here. Maybe we could go to the next slide. Building your Agentic Enterprise today, I think, requires a couple of things. You know, one is having that trusted team. And thank you. Thank you to my team. And that's all of you, the Trailblazers, the community, the partners. You need forward-deployed engineers. You need people shoulder to shoulder with you. We're your team. You're our team. Our professional services team is here for you. Success plans, our partners, our ISVs, and a trusted team can deliver trusted AI. But, you know, we need something else. We need also a trusted agreement. We need a trusted way to link value of AI with our pricing and our packaging. And that's why we have these new Agentic-Enterforce agreements that have unlimited Agentforce and unlimited Data360, because we want you to have trust in that process as well. And we want you to be able to build and grow with the Agentforce ecosystem. It's not just starting with our agents, but so many agents from our partners are out here today. I hope you'll visit many of them. The Agentic enterprise journey at World Tour DC is just beginning. Hope you had an amazing keynote, but I want you to see all of these sessions. Patrick's got a session. So many of our key technical leaders have a session. Check out the demos. Go to these breakouts. Thank you to our partners Accenture, Kairosoft, Deloitte, IBM, and I missed somebody else, but thank you, everybody. Please take the survey and give us feedback. We love you. Thank you, Washington, D.C. Thank you, the DMV. Thank you to our veterans and our service members. And all our public servants for everything you're doing. We love you. Have a great day.