Win and Scale Sales with Contextual Data | Data 360 Keynote, Dreamforce 2026

Hello. So for the last few days and for several days, all day, you've been hearing that data is the foundation for the agentic enterprise. However, more data won't save you. A smarter model will not either. For AI to be converted to business outcomes, what you need is context. Trusted enterprise context. That is your AI advantage. Hello, Dreamforce. Good afternoon. Good evening to folks in the room here and people listening online. So I'm delighted to be here. Before we get started, I want to say a thank you. Thank you to our employees. Thank you to our partners, our customers, our Data Blazers, Trailblazers, and the rest of the community. Over the last 3 years, we have gone through a transformative change with data and AI, and we have journeyed together with that. And thank you for putting the trust in us with your data and your customers. We are going to journey together again going forward. The best is ahead of us. And thank you for the incredible growth that we've been seeing. Since the last 12 months I've been in front of you, Data360 has grown incredibly well. What you have told us is that enterprises have trapped silos of data, and we unlocked the trapped data to deliver business outcomes. And we have grown incredibly over the last 12 years. I can throw a lot of numbers at you, trillions and quadrillions, but what is astonishing is the growth— 200% growth on a strong foundation. And over the last 12 months, we have seen incredible growth with unstructured data. That is from agents now using Data360. Now, here's an interesting stat. That 263 trillion rows of data that you've seen, records that you're seeing, 70% of that came from zero copy. This is the zero copy network that we work with our partners to invent. What we're doing here is not moving data from one silo to another silo. Instead, we are zero-copying it and we are making data fluid. Data fluidity is where we are driving toward driving the customer outcomes. It's not just about scale, however. It is about business outcomes. Yesterday, I was with the leadership team from Akamai. They started by telling me that they're results-focused, they're outcome-focused. So if you take a look at the logos here, whether you're looking at the biggest of big customers like FedEx, driving 2,000% ROI, or a small customer like Mimit Health. The CEO of Mimit Health, Dr. Chopra, is also their surgeon, is here in the room. A small company, a small provider network providing business outcomes and better CSAT. Now, along the way, you've also been telling us and giving us feedback. This is what we've heard from you. Time to value matters. Whether you're delivering customer service through a contact center with a real-time data graph, or whether you're doing a marketing campaign with a segment with harmonized unified data. How quickly you get set up, how quickly you unlock the trapped data, and how quickly you deliver business outcomes is paramount for you. Now, with agents, what we're hearing is that with a probabilistic and and deterministic outcomes that we have now pioneered at Salesforce. With probabilistic outcomes, accuracy matters. And not to mention the runaway costs with LLMs. So what we are looking for is cost-effective smart agents. And AI where you work, whether it is the interface of your choice or the harness that you use, what you're looking for is AI working in the place where you work and being able to deliver value through the agent persona. Now, what's common across all of these? Trusted enterprise context. What you need is trusted enterprise context. Now, let's unpack that. So on Monday, I was having a discussion with Michael Marr, SVP of Technology, for Wyndham. What Michael told me is they've been customers of Data360 for, for several years. One of the stats he shared with me, business outcomes, was that they got 300 basis points better conversion through their engagement with customers and 15% improvement in daily rates. Now, how did that come about? It's higher quality engagement and better conversion rates. It's about providing preferences and relevant rates to the engagement. That is context. That's contextual information provided to agents, humans, APIs, and apps in the place where they're engaging with their customers. Now, context that they're using is with the right semantics, right knowledge, right permissions, and also right activity and memory. That is trusted enterprise context. Now, just because they had access to all of that data across their entire enterprise, it does not mean that that is context. Context needs to be assembled. It needs to be governed so it becomes trusted. It needs to source through the entire enterprise so it's trusted enterprise context. Needs to be remembered, needs to be reused. And that's what Windham's been doing. And also, the breakthrough that we have here is not about providing the context to every app And every agent, every API, or every application, it is providing the minimum context needed for the maximum result with optimized cost. That is what our customers have been doing now, and they're seeing remarkable amount of ROI. Let's take a look with what our customers are doing. Let's roll the tape.

Data is everything for an agentic enterprise.

We have so much data.

But how are we really using that in a way that moves our goals forward? Patients are so unique, and that makes data and technology problems very hard to solve.

We wanted to bring in data so we could start to talk to our fan base and understand who they were.

We talked ourselves into a narrative of this will be impossible. The reality was the right data to power our agents was available in Data360. Agentic AI is only as powerful as the systems that you connect it to. Making sure it's connected to the right data is mission critical.

Data360 is at the heart of our infrastructure. Without it, we wouldn't be unified.

I could not have turned those agents on so quickly if I didn't have those Data360 capabilities.

These agents have a huge impact to that fan experience.

Customers are happier, we're getting them answers faster. What's the trade-off? There's no trade-off. With the service agent, we saw a 60% improvement in deflection rate.

We saw a 40% reduction in held calls and 30% reduction in calls for support.

We've reduced our case handling time by 45%. Pre-procedure prep for doctors went from 90 minutes to 90 seconds. That's the power of this. This agentic enterprise is the Iron Man suit.

We're dying to see where this will take us.

When we're working with Salesforce, it feels like we're right at the front of the pack. So by having Data360 sitting behind our agent, it just makes it all possible.

Isn't that incredible? So that is what Data360 can do— data at scale and context at scale. You saw this architecture from Mark. yesterday, Data360 is at the, at the center of everything. It's a foundation for everything. Now let's see how Data360 is activating trusted enterprise context. Now, you— we have been the number one CDP for the last 3 years thanks to all the efforts that we've had from customers, from our analysts, and the many analysts here in the room. Thank you for the recognition. Now, on top of being the number one CDP, we have innovated, and now we have our first agentic customer data platform that's GA now. Our customers are using it. Now, I refer to the 800% growth in unstructured data. We have improved on that innovation over the last 12 months, and I'm delighted to announce today that we are releasing Agent Context Engine. We're going to dig deeper into this. And then finally, we're looking at Agentforce Coworker. You've heard about Coworker everywhere. Mark talked about it yesterday. You've seen demos. Agentforce Coworker is a canonical representation of a real live agent that sits inside Salesforce in the Lightning interface and other places where you can access Salesforce engine through a headless capability. And it leverages the power of the trusted enterprise context. That is GA now. It's been GA for 35 days, and we have seen incredible growth. 110,000 users gone viral in the last 35 days. We're going to dig into each one of these. I want us to get started. And to start— kick us off, I want to invite James Nakashima to the stage here. James runs our product team, and James is going to walk through our number 1 agent, xCDP. James, take it away.

Yeah, thanks, Rahul. It is so great to be here at this transformative time. I mean, AI is everywhere, but it's interesting. Even though AI is everywhere, so many of you are telling me that you're still using yesterday's tools to find an insight in your data, that you're swivel chairing between apps, agents, and the data that actually drives your business. Well, the good news is you don't need to anymore. Data 360 changed the game when it came to customer data, and we're going to change the game again. Data360 is now headless and agentic. Other CDPs, they're adding chat. We did that last year. We are combining and unleashing the power of agents and the world's number one CDP everywhere you work, thanks to APIs an MCP server, and skills. Okay, but let's look at this in action. To help with the demo, I want to invite April Moon, one of our resident Data 360 experts.

Welcome, April.

I got some fans. All right, James, I'm ready to go. I've got my Data 360 instance has been provisioned, and I'm ready to start building on it to make an Agentic CDP out of it.

Awesome. And you're using Claude, but it could have been ChatGPT, could have been Slack. Any MCP client. How is it that these, you know, chat tools actually know how to operate Data 360?

Well, most chat tools are connected to anything that's public. All of our documentation is public, so you guys all have that information. But mine is actually also connected through the connectors to my Data 360 instance. So it knows based on my permissions exactly what's in the org. And the first step to do on any of these deployments is to see what's already in the org and what's connected, what's not, How many segments do we have? How many activations are running so far? And we don't have that much. So we've got 2 mapped data sources, still got 14 to go, no segments, no activations, but individual and account are ready.

Okay, so not starting from scratch, but a long way to go. It's great though. Claude's saying that there's a plan.

Exactly. These chat agents are great 'cause they're looking at it and they know that the end goal is to be able to market out to our customers and to create campaigns. So to get from our 2 connected sources into an actual campaign, we need to go ahead and kick off mapping and then identity resolution. And because the tools inside the Data 360 MCP are there, and Claude in this case already knows they're there, it can map all that information for us based on my instruction and come back with our commerce and loyalty mapped in to add to our account and individual. And then once our customer identity resolution is all done, we've got almost 3 million unified profiles ready to go.

Wow.

So that's awesome. One of the things with other CDPs that often is very costly, time-consuming, even though it's one of the most important things that a CDP offers in terms of deterministic value.

Exactly. And, you know, we did all this headless. But you did say agentic CDP. Agentic does not mean headless. I can also do things inside the org all on my own, and we are building some pretty cool stuff out. So you all are going to get a sneak preview. You'll get this coming in the new year, but we have a new segment builder, and this is all an agentic experience powered in here. And I can just describe exactly what goal I want and kick this off. And the Data 360 agent is going to go through, it's going to understand what's already deployed on my org, understand what instruction I've given it, make recommendations that I can choose to accept or modify, and then it's going to return to me some ideas around what kinds of segments that I need to create based on my goal. I kind of like this one that's a little bit more narrow, narrow, and we're going to go ahead and build it. Now, I have built out many a segment in my day and I've hunt and pecked for the right rules and the right fields to do. My agent here has done all of that for me. So I have it and I can just accept, save, and be on to the next project.

Yeah, so fast and easy. Data 360 was helping you every step of the way. I noticed also that you hinged on loyalty. How do you know? How do you have confidence that that's the right field?

Well, one thing I've definitely learned is to not trust the AI 100%. And we've got you covered there as well. So with Data 360 Lineage, I can find that exact loyalty field that it was pulled on, and I can trace it all the way back to the point of ingestion, all the way forward through other segments and data model objects.

Awesome. This lineage view is super helpful to just understand kind of origin all the way to multiple destinations, activations, what's happening with your data. Exactly.

Yeah.

Hey, thank you, April, for the wonderful demo.

Thank you.

So that's the power of Data 360. the world's number one CDP, and the power of agents combined, available everywhere you work. But it's more than just that power. It's— whoops, back please. Oh no, sorry. We got it. 200 innovation. It's about what our customers and analysts are saying about us. Forrester, Gartner, G2, IDC, they've all said the same thing, that we're the leader. But what matters more to us is that you, our customers, see us as the world's number one CDP. Here's one of the reasons why our customers tell us we're the number one CDP. We've been delivering value quarter over quarter, over 200 innovations in the last quarter. Let me just talk about a subsample of those, expanding our zero-copy network, and Private Connect, Salesforce Catalog for metadata, unifying to households and to an MDM defined in Informatica, adding predictions and enrichments to your data on things like buying behavior and sentiment, and then expanding the reach of your ad audiences with Identity Boost. That's one of the reasons why our customers say that we're the number one CDP. But maybe the most important reason is our vision and strategy, something that's been playing out for a number of years now. It starts back years ago when it was a CDP for marketing, and we expanded that. We expanded that to Customer 360 and to agents. Data 360 is the investment that you can make today that accrues forward as you build out your entire agentic enterprise. So I'm going to flip it. I've been talking about how agents make Data360 better, but let's talk about how Data360 makes agents better. To talk about our context engine, I'm going to invite one of our most respected engineering leaders, MK.

Thank you, Jamie.

Yeah, thank you.

Proud to represent engineering here with our famous shoe there. All right, so what you saw so far was, like James said, how does Data Cloud help all agents and how are we gonna sort of use agents to make data better? Now, we are all gathered here. We are all a great team, aren't we?

Yeah.

Yes, sir. What do great teams do? You're all here to learn, to remember, to share, and work together. That's what makes great teams kick. But when you have a lot of agents, what happens? We solved all the data silos, but we just end up creating more agent silos because every agent is going to have its own memory, its own context. You're going to have different agents running, all thinking its own thing. And why? Because agents need 5 different things. One, certainly they need the data. What is true across the enterprise? What's the right thing? They also need to know how to efficiently access that data. This is where things like RAG and Graph RAG and other things come on. They also need to know the semantics of that data. What does that data actually mean? More importantly, they need to know all the memory of the conversations you have with the agent. and know who to share that memory with and the learnings that come from there. And that's where governance comes in as well. These 5 things are what can make an agent a great team as well. Now, enough said with that. Let's actually see how all this comes together with our agent context engine. This is how You're going to be able to take all the data that James showed, convert it into real-time trusted enterprise context, and share across all agents— agent force agents and non-agent force agents as well. And remember what Rahul said— dynamic context, which means only share what's needed so that the agent can make the right sort of reasoning with the least amount of cost. April, are you ready to show the art of the possible?

Yes.

All right. Now, whatever you see here with Rachel, it's all production and real code. Because once I asked her, hey, it looks all fancy. Is it real? She got really upset. She's like, I code. Everything is real. So yes.

I built this.

Exactly.

So just because, you know, we interact with agents all the time, I am a racing fan. I built a little skin so you guys can see the phone. But, you know, when you ask an agent on race day, maybe, What's going on? Where are we at? Where should I go? I'm logged in, so the agent knows who I am, but the agent doesn't really know where my tickets are. It doesn't know where I'm physically at. And it's kind of a little bit of a boring answer. Would we say that?

Why is that? Because it doesn't have—

Because it doesn't have context and memory.

Exactly.

So we can add a little magic.

Look at that magic wand. Hey, but I only see a few things dropping there. What happened to that? Unified proof.

Well, I'm authenticated, so the agent didn't need to re-understand who I am in order to answer the new question. It already knew that. So all it did was package up the new context that it would need to answer my question appropriately.

Right. And that is dynamic context. Only pick the things that you need to answer the question.

Exactly. So if I continue asking questions about maybe getting a little bit closer to the pit lane, It's going to change it up again. It's going to understand some different things and it's going to pull from some different information that it's all connected to.

And what was that purple thing that ran this side?

So that's the memory. The agent understood because I just said, hey, can I get closer to the pit lane? It's like, hey, maybe I need to remember that later. And it created its own learning from what I gave it and it wrote it so that it could use it later.

Right. So this is important. Memory is not the same as facts, right? So memory is all the interactions you have, you kind of remember, but there are things that you distill, you learn from those memories and you create facts. Those facts become part of the profile.

But those facts aren't just locked inside of one agent. So let's say I'm kicking it on my couch the next day at 2 in the morning watching the races. I can log in on the website.

That's a different agent, right? That's on the website. We need a different agent.

Different website. It still knows who I am. I'm still authenticated. It's still connected to the same context. And that memory from the previous agent is available and there because they are all connected. So when I continue my conversation, I don't have to continue to restate everything that I'm doing. I can check in for my virtual experience. It's going to understand this information and then it's going to give me a wonderful answer.

Amazing. So no more agent silos there.

No more agent silos. But MK, you did say teams. I work with agents, but I also work with—

Humans.

Yeah.

Humans.

All right.

So all of those exact same memories are available for, say, a human answering a phone and using the service console.

That's the service console that you see, right? And you see all the things on the right? That's the context. The same thing the agents got, the humans also get. So it's not just great teams of humans, great team of agents, it's great team of humans and agents working together.

That's it.

Thank you, April. Back to the slides. So what you saw was one context working for any agent, any app, any data. You take all your data, structured or unstructured, A lot we can bring with zero copy. And what you saw was real time. You saw how these things were able to talk across agents and humans in real time and sophisticated techniques like GraphRAG to be able to go understand all of those unstructured data. That is what that agent context engine is going to give to you on top of that data layer. And more importantly, that context can be shared across all agents that you have in your enterprise. Now, speaking of context, every time I have some home issues and all that kind of stuff, I go and run and get the best deal possible at the best home improvement store that you're all familiar with, which is Lowe's. I would love to invite customers here, Abhishek and Amanda from Lowe's, to tell us their experience with how they're using Data360 and the Cortex Engine.

Abhishek, Amanda.

First question to you, Abhishek. Lowe's took a very different approach to Data360. Instead of trying to say, you know what, build one customer context at a time for every scenario, you took to say, okay, let me create one customer profile and then go build many scenarios out of it, right? So why do you think that was important and what can you sort of tell us about your experience and journey?

So Lowe's is an omnichannel retailer, but for customers, they don't think in terms of channels, right? Whether they're in the store, on the Lowe's website, or interacting with an email or an SMS that we send, for them, it's Lowe's. So for us to be truly relevant and contextual in those moments, we need the full picture of the customer. Lowe's uses Google BigQuery for all its data needs, right? That's where our own Customer 360 sits. and all the golden tables, right? From there on, we move the data into Data360. Now imagine this, if a marketer did not have access to an attribute or a trait, right, how would they even activate on it? How would we make the experience contextual, right? That's the only reason we can do it.

So you had data in CRM, data in Google BigQuery, you're bringing it all together for that marketing.

Absolutely.

Right, awesome, thank you. Amanda, turning to you, so sort of talking about the marketing, You put all that context to work for Milo's Rewards, right? How has your marketing team and the experience changed using Data360?

Yeah, it's changed in a big way. One of the things I'll say is the unlock goes far beyond using the data to drive personalization. You know, we run a loyalty program and now we're not just running a program, we're really running an enterprise growth engine. And what I mean by that is previously, you know, a program or a piece of data can personalize a reward or a moment in time. And now with the context and the data that we have access to, we're answering a much bigger question. We're starting to understand of everything we know about this customer, how do we create value for them? And then how does that value deliver back to the business?

Right.

And so when I think about the power of this data foundation, it is a growth lever for us. You know, it really allows us to place intelligence, as Abhishek said, not just in a single channel or a single moment, It's taking that and it's the foundation of our whole relationship that powers this growth flywheel. So as we get a greater customer understanding, that turns into greater relevance.

Yeah.

And then more relevance turns into more value for the customer.

Right.

And that more value drives the business.

Yeah.

And so that's really what we've been after and truly unlocking. And now this customer understanding is effectively a strategic asset that grows over time, and it's allowing our business to grow, the brands that invest in our media network to access it. And we're only just starting, and it's been a big unlock for us.

That's amazing. I mean, I love those personalized emails that I get, right? It's like perfect. I hope so.

I hope you open every one.

Yes. Now, going from humans to agents, right? You've sort of built this audience agent, right? And now you're trying to use this context beyond that to go into these sort of agentic experiences. So how do you see this human and agent sort of working together with all this context?

Okay. So I talked about the full picture of the customer, right? And when we say full picture, I'm talking about 700+ traits and behavior, right? That's what we have in there. Imagine asking a marketeer or any operator to remember all of that and always figure out when to use what, right? That becomes a daunting task. That's a good use case for agents. What we did was we created the business context, the domain knowledge, That's our hardness, right? And we allowed our operators, the marketeers, and anyone that uses this data to use natural language to access it, thereby making all of that data accessible. Now that's used for our digital experiences, our marketing use cases, our media network, all of that. So having the data is one thing, making it accessible is really, really important.

Right.

Wonderful.

Thank you. Thank you, Amanda. Thank you, Abhishek. and go sign up for Milo's rewards. And so all the emails you're going to get, all those personalized experiences you get, thanks to the two of them, as well as what you see in Data360. Now, we saw the agentic CDP. You saw sort of the ACE context engine. There is one out-of-the-box agent that we have built with Agentforce that's going to make all of this power just show up in all of your orgs today. To talk about that, I would love to invite Rekha.

Thank you, MK. Thank you all for being here. We are very close to getting to Dreamfest. We promise it's going to be a great show. But before we get there, I want to take a step back to talk about enterprise work as it stands today. There is not a particular tool, not a particular agent, or not one team that we all work in, right? Our work spans Agents, apps, tools, humans, teams, you name it. That is exactly where the enterprise work is broken. So what if you didn't have to do that? What if you didn't have to toggle? What if you had everything you need to do right in one place? So excited to introduce you to Agentforce Coworker. This is your autonomous AI teammate, lives right inside Salesforce, and gives you everything you need to get work done. done. And remember, this is built on Data 360. It's built for the enterprise, which means it's trusted, it's secure, it's governed. It has all of your context, your data, your metadata, your workflows, your permissions, you name it. It has everything you need. And it's not just inside Salesforce Lightning. It's across cloud. It's across Teams. It's across ChatGPT. you have it everywhere. It's that one coworker that you will want to take everywhere with you. Enough of me talking. I would love to show this in action. To do that, please welcome April again. So she deserves it.

She's amazing. We did say it was kind of a pre-show.

So every morning I'm on my phone, And I pull up this mobile digest by Agentforce Coworker. I can quickly scan to see what's working, what I need to focus on today. It gives me everything. Looks like I have no big fire drills going on, so I'm good to go. I'm going to go get ready. I'm going to get my kids ready to school. And with that, I'm going to come back to my desk to see what's up with Salesforce. Our home screen, this is where I start my day. And right there, that Ask button is how you access Agentforce Coworker. You click on it. So April, let's ask Coworker what I need to focus on for today.

Sure thing. You can be as specific as you want. The more specific, the better, as always when working with agents, right?

Absolutely.

And it's going to go ahead and turn through— it's going to pull through everything that Rekha touches across this org, and it's come up with a great summary for her so that she has everything as she finishes her morning coffee.

Okay, this is way simpler than how we would have to toggle multiple apps to get this done, so this is cool. So, notice how it gives me my top accounts, my top deals I need to focus on. It also shows that I have an upcoming dinner meeting with United Oil tomorrow. I really need this deal, and Allison is watching me, so I'm going to try to close this as fast as I can. So, April, can you help me prep for this call?

Of course.

Let's just go ahead and set Coworker off to help you prepare for the meeting. And it's going to turn through everything it knows about United Oil, flag anything that you need to know so that you can walk in confidently without your laptop.

Can you imagine us working without Coworker? Like, all of this prep would have easily taken us, like, what, 2 hours to do? This is pretty cool. But notice how Coworker is pulling up the people that will be in the room for my dinner. It also shows that there is an open case. How many times we've been with customers where the customer tells you, hey, you know what, I have an open case with you. Can you go tell me when that needs to be fixed? I can swear that I would not have seen that when I'm prepping for the call, but Coworker's got that for you. So Coworker here is quickly alerting me that there is an open dispute case, so I need to make sure that this gets fixed. And think about the customer experience because you're now walking into a customer meeting knowing everything about the customer, their open cases, their upcoming opportunities. So you become instantly the trusted advisor. And here I'm going to ask Coworker to quickly draft up an email to send to the customer before the dinner meeting so she knows we are on it. Let's do that, April.

Sure thing. Super quick and easy, and it is ready for you to review.

Okay, this is awesome. I love that, and I'm going to get this to her. So she's going to be happy knowing that we are working on it, and the dinner meeting is going to be great. Wish me luck. But while I have you, can you quickly toggle to the accounts, to the opportunities? I had this call with Blue Water last night. They were very happy and they gave me a verbal commit, so I want to close it. While I have you, can you just quickly toggle?

No, thank you.

Oh no, what are you going to do?

I'm just going to ask Coworker to go ahead and change that over to commit in your forecast for you. And it'll scan through, it's going to understand based on the context that I gave it exactly which deal that you are talking about. It's going to give me a way to confirm that. And without leaving the space, we've confirmed it and we are committed.

Oh, wow. So what you just saw in this screen, I asked Copilot what I needed to focus on today, gave me my top deals. It also helped me prep for the dinner meeting tomorrow, gave me a proactive email I can send, and closed the account. All in one screen. We didn't leave the screen. And also notice how April did not have to call the specific agent. She did not say anything about the sales agent or the service agent or marketing agent. None. She just explained in plain English and coworker did that for us. That is magic, and that is what AI should be. April, thank you. And before we go, hold on. If I show this, just like all of you who are excited to try Coworker, United Oil is going to want to do this too. Unfortunately, they are a Teams shop. So wait, what?

We've got you covered. Because we're headless first, we are extensible, and I already asked that same question earlier today in Teams, and I have the exact same answer ready to go.

Amazing. Now you can clap. Thank you, April. You know, Coworker has been truly magic, right? And you saw it, you see the value. I can assure you this is going to be a game changer for you. But this is actually my most favorite part about Coworker. It's so fast. You can get it up and running in literally minutes. You can enable Coworker in the Salesforce setup. It's a quick toggle. connect all your data, and grant that governed access to your users, to your agents, you name it, and you have it ready to go. So scan that QR code so you all can try it on today. And just like us, we have another huge fan of Copilot amidst us. And to do that, I would love to welcome Stephanie from Accenture. Stephanie, thank you so much for joining us today. I am really excited for this partnership. You guys came up quickly with Coworker in the last month or so?

Yes, the last 35 days.

Incredible. How has your experience been with Coworker?

Yeah, it's been great. So, at Accenture, our guiding philosophy is we have to go first in order to help guide our clients, and there is no substitution for real-world execution. And so, when, as we've been working with Salesforce on this very quickly, We have been building agents since 2024.

Wow.

And one of the things we noticed is our users, they don't want to take the time to figure out which agent to go to. And so one of the feedback we kept giving you is, how do we get better about this? Enter Coworker. And so we brought Coworker into our environment, and overnight we deployed it out to 16,000 users.

Wait, did you say 16,000?

16,000. So of your 110,000, We're 16,000 so far and growing. And it was amazing because we did it very silently to see what would happen. And overnight, our usage jumped an order of magnitude.

Wow.

Because what it did was it took away the need for the end user to figure out which agent to go to, and they could just ask questions in their business intent, and all of a sudden, it was serving it up. So exactly what we just saw in the demo is what we have within Accenture now.

Oh, I love that. And you've been great partners to us, so appreciate you going as fast as we can. We also know that you're working across— your teams are working across Salesforce teams and other AI tools. So how do you bring that trusted context everywhere? And what would you say are your early signals that you're seeing from Coworker?

Yeah, so the trusted context is an important one. When we started building out the agents, we had to go through a lot of the different 5 steps that you highlighted earlier and getting all the data right, getting the governance right, getting all those. Once we had done that, we could turn Coworker on, and all of a sudden, now the context was there. Business users could ask exactly what they needed, and they could get what they need to run their business efficiently. As you mentioned Teams, we are a big Teams shop, and we also start to think about who are the personas within our company, and what is the right user interface for them, and how do we serve this information up consistently. And so for some people, it's going into Salesforce because they operate there every day. But for a lot of other people who are just like, I need to move this deal close from the end of this month to the end of next month, they just want to do it in Teams. And so going into a tool that they're already in, being able to type this one sentence and have it go out was beautiful. And again, we also turned this on for 16,000 of our users overnight quietly. And people just found it. And the beauty of when things work is then everyone starts talking, and they're like, hey, did you know you could do this? And again, our usage just spiked by an order of magnitude. So it's amazing.

Oh, I love that. Thank you so much for joining us today. And I hope you have a great DreamFest and the rest of Dreamforce.

Thank you.

Huge round of applause, please. So that is Coworker, your context, Every Surface, one agent-first coworker. And with that, I'm going to pass it back to our incredible GM, Rahul, to wrap us up.

Great. Thank you. Give it up for Rekha, April, Murali, James. So what have you seen so far? Remember we talked about time to value? That coworker thing literally can light up within minutes. And James showed you the ability for you to drive segments. and activate them with the ease that we didn't have before that we have now innovated with the agentic customer data platform. And we also looked at activation with our ACE, Agent Context Engine. Now folks, remember, we are in the home stretch here. I just want to tell you that everything that we saw there was live code. It's production code. And as a matter of fact, when April was showing the coworker demo and she was showing the CDP as well as the ACE demo, She didn't even pre-curate any of that. She literally did it live in the moment. So we also looked at Coworker in terms of its speed of growth. Stephanie, thank you. I know you are part of the 110,000, you're 16,000 in there, and we are growing rapidly there. So technology makes things possible. Our community makes things better, and our community has been driving. We had a Datablazor community that we announced 2 years back. Now it has grown remarkably, tens of thousands of people joining us. every year. Now, these Data Blazer community has been an inspiration. To bring it to life, I want to bring one of our incredible Data Blazers who's here with us in the audience, Angelica Bufa. Do you want to join us up here? So thank you for being here. So Angelica inspires us. She inspires me. She inspires the team. She inspires the Data Blazers. So you've been doing this for 15 years with data. What have you seen things have changed over a period of time? Give us your perspective.

When I started my career, we were thinking about data, you know, as something that we can store, transform, and move around, and of course report. And we used to spend a lot of time, you know, on the technical implementation of the data pipeline. And the biggest change is the shift in how we can perceive data. We went from thinking that it's just another technical architect— artifact, sorry, to actually being something that has become an important part on achieving trusted context. And Data360 is also playing an important role right here because it's what we use to define that data foundation that then we can reuse and evolve as, you know, our business requirements change. And that is one of the first steps to actually building that trusted context. And what really excites me is the fact that then we can hand over that context to AI and agents. So our teams can stop thinking about the technical things and answer more interesting questions like, what are we gonna be doing next? So we stop being stuck in operations, like dealing with all the details, technical details, and actually moving forward to what we're gonna be doing next and actually activating that context.

That's great. So you've been at Dreamforce for the last couple days, and we've been engaged with you on an ongoing basis. You've seen what's coming. What gets you excited most? What's coming?

The fact that now we're going to be handing that information to our agents, and we're going to be capable of leveraging that trusted context. And again, it's that shift from going to, oh, I need to worry about data that we're going to integrate, and Now I can just forget that and move into more interesting things. The things that are like actually giving me the most return of my investment. It's like what we're going to be doing next, how we can interact with our customer, how we can activate that.

That's great. So like I said, you have been an inspiration for us. You have been an inspiration to the community and to our customers and partners. For all the incredible work that you do, we have a token of appreciation for you. We want to hand you a golden hoodie.

Yeah!

So Angelica is not the only winner here. We have additional winners. We have a golden ticket under 15 chairs in this room. Reach out below your chair and look for your golden ticket. You're winners too. We have $250 in goodies under your chairs. So looks like some of you found it. Congratulations. So to all the winners here, congratulations to Angelica and to the winners as well. Thank you for the inspiration. Your journey doesn't stop here. Your journey continues at Dreamforce. And just one more slide here. Can you get back to the slides?

Yes.

So your journey doesn't stop here. It continues at Dreamforce. We have several sessions planned for you. Go down to the campground. We do have the DataSeeds 360 Lounge. We have experts available. We have theater sessions. Look at the barcode. The barcode tells you everything that you need to do for the rest of your stay here at Dreamforce. So we ended this keynote Talking about cost and efficiency, 4 minutes and 33 seconds ahead of time. So there you go. Thank you. Thank you for being here. Thank you.