Dreamforce 2024 | David Schmaier | Salesforce

Clips
News
More from Dreamforce 2024

David Schmaier

President & Chief Product Officer

Salesforce

Salesforce reports mixed third-quarter results, but stock jumps on upbeat forecast

Shares in Salesforce Inc. rose more than 9% in after-hours trading today after the cloud-based software company impressed investors with an upbeat outlook despite reporting mixed results in its fiscal 2025 third quarter.For the quarter that ended on Oct. 31, Salesforce reported adjusted earnings of $2.41, up from $2.11 in the same quarter of fiscal 2024, on revenue of $9.44 billion, up 8% year-over-year. Earnings fell short of the $2.44 expected by analysts, while revenue came in ahead of an expected $9.35 billion.Salesforce saw subscription and support revenue in the quarter of $8.879 billion, up 9% year-over-year, with an operating cash flow of $1.98 billion, up 29% year-over-year

How Salesforce Data Cloud is powering AI-driven customer experiences

As we look at the current trajectory of enterprise technology, it’s clear that platforms such as Salesforce Data Cloud are essential for driving artificial intelligence-powered customer experiences.Organizations are pushing toward more intelligent, human-like interactions across customer experiences, moving beyond chatbots to AI-powered agents. This trend reflects a growing need to leverage both human expertise and AI capabilities to create seamless, scalable systems that redefine customer relations, according to David Schmaier (pictured), president and chief product officer of Salesforce Inc

play_circle_outline Unlocking Success: Agentforce as Central Focus in Fostering Collaboration Between Humans and AI Agents

play_circle_outline Data Cloud for data normalization and integration

play_circle_outline AppExchange for building and using agents

play_circle_outline Ecosystem-friendly approach to agent building

play_circle_outline Transition from application company to hybrid application and cloud platform company

play_circle_outline CEOs interested in AI for competitive advantage, disruption in various industries.

play_circle_outline Agentforce aims to bring disruption and ease of use to companies.

play_circle_outline Value captured in organizing expertise, fine-tuning models, and AI in industry-specific work.

Info

David Schmaier | Salesforce

David Schmaier

President & Chief Product Officer Salesforce

search

>> Hello, welcome back to Dreamforce 2024, live from the NYSE, where we have set up just outside of Moscone to really bring to you what's going on and what you should expect. This week we're really excited, because, again, we get some of the best guests that Salesforce has been really great about bringing on here. And we have David Schmaier, who is the president and chief product officer with Salesforce, who's going to help us break it down. I'm also joined by George Gilbert, who's our data guru-in-chief here, and we're really going to dig in on some of the announcements that's been going on. I think one of the big things that was talked about last week was really Agentforce and the announcements around that. Help us understand where you're thinking and how to get... People always talk about bots and getting past bots, so how does this announcement really help with that?

David Schmaier

>> Sure. And by the way, thank you for having me here today, really appreciate it. Great to see you both. If there's three things that you take away from Dreamforce this year, it's Agentforce, Agentforce, and Agentforce. So we're going all in on humans and AI agents working together to help customers, and we really think this is a big story because a lot of companies have been trying to build their own AI, and we don't think this is something you want to do it yourself. If you go to a surgeon, you don't do this yourself, you go to somebody who really knows what they're doing. And what we've done is built this out-of-the-box so you can literally turn our agents on in hours, and not take months or even years to build the AI yourself.

>> So let me just ask at the top-level this value proposition, there's always this tension between an application that is out-of-the-box value that's packaged and the desire to allow it to be building blocks for custom applications, and no one's ever really been able to make that both a dessert topping and a floor wax, to borrow a phrase. Agents seem to get us closer to that. What is it about agents that allow you to provide the best of both, and what did you have to do to the foundational components of Salesforce to enable that?

David Schmaier

>> That's a great question, George. We really have built Agentforce on top of the foundation we built over 25 years, so multi-tenant, in the cloud, proven, scalable, trusted. On top of that, we built something new in the last several years called our Data Cloud, which allows you to basically integrate and harmonize and unify all the data together. And our view is the data powers the AI and the agents, and the agents power the customer experience for the next generation. So we've really brought it all together with Agentforce, and in my opinion, I've been working on what we call CRM software for over three decades, this is the dream of CRM, where you and I, the three of us all work together and we put data into the system, and the system magically does things for us. Now, the AI agents can do that magic.

>> It would seem that normalization of that data is a big task, and again, having been and used Salesforce in multiple different ways in multiple different parts of the cloud, Salesforce Cloud, how did you really achieve that helps people normalize that data so that AI can really be actionable, really be trusted, within that?

David Schmaier

>> Sure. We've been working on AI for over a decade at Salesforce. We have a Stanford professor of AI and machine learning who runs our research center. So we've been working on this for literal years, and we do over two trillion AI transactions every week. But when GenAI came out and ChatGPT came out, the whole industry changed. And so, what we've found is this is a real opportunity to double down and go all in on agents and AI in the flow work, so that's what we've done, and what I'm most excited about are the customers. So customers like OpenTable and Disney and ADP and Wiley and Saks Fifth Avenue, you're going to hear from them this week, you're going to hear their stories, they're going to talk about how the agents have literally transformed and revolutionized the way they interact with their customers, so they can do what has really been our vision for 25 years, that we help companies connect with their customers in entirely new ways, and Agentforce is definitely an entirely new way.

>> So Data Cloud, it took a few years to pull together a historical system of truth, and even some harmonized integration to the real-time system of truth in the operational applications. But what seems to tie it all together is you have a common language for Customer 360 and the customer journey, that was critical, and none of the other data clouds really have that. But then, you also have the common language for the business processes, which now become the tools or actions that the agents can invoke to get work done. Now, as far as we understand, no one else has all those ingredients, because you needed those harmonized actions not just to get stuff done, but by connecting it to the applications, they can learn from the outcomes, and so the agents get smarter over time.

David Schmaier

>> That's exactly right. I think that's very, very well said. So we have always had the full Customer 360 view, and some of the other CRM players give you what what we call Customer 360, they give you a customer 90 or Customer 180, only Salesforce gives you the end-to-end view across every single channel of communication, across sales, marketing, service, and commerce. But now, the data lake, the Data Cloud, takes it further, because we can bring in SAP data, we can bring in billing data. If somebody clicks on a certain product page on your website, you instantly get that. That's a signal that we can send to the right salespeople and say, "Hey, somebody just clicked on this page, they might want to buy this particular product." That would be good to know instantaneously, because leads are sometimes perishable. So we've connected all this together. The amazing thing about Agentforce is we have over 200,000 customers at Salesforce, and it's all built on these core fundamental technologies, like at the data layer and at the workflow layer, our flow automation layer does 44 billion workflows a day, and so it's used and beloved by all of our customers. So if you use Agentforce, now with a little bit of upskilling... If you know how to use Salesforce, pardon me, now you know how to use Agentforce. And so, we did a training yesterday of our own people who we're going to turn AI agents on for people right at Dreamforce, and we had our own people turning on agents in one to two hours, working, really smart, intelligent, conversational agents in one to two hours.

>> You said something interesting, where you make it so easy to harmonize all the customer data in one language, Customer 360, the journey, and that might make it easy, there's almost like a network effect because it draws in non-customer data that's related to the customer. So is that one way of bringing, turning this, what was a customer system of truth, into a broader system of truth, and then through connectors, interacting with other operational applications? In other words, is the aspiration to become the application platform for agents beyond just the customer surface points?

David Schmaier

>> Sure. Well, the data lake, our Data Cloud, allows you to... We now have over 400 connectors, and with MuleSoft, which is the number one API management system, we can get you to thousands and thousands of applications. So we can virtually connect to anything, but we can also connect to other data lakes, like Snowflake and Databricks and Redshift and BigQuery and Azure Fabric data, so we can connect to any of these other data lakes. And we just announced a big partnership with Workday and with IBM and other big partners that are also joining what we call the Zero Copy Alliance, where you don't need to copy all the data into our Data Cloud, we can just point to it, but give you a logical view of that. So what you said, George, is exactly right, that what we used to give you, the end-to-end view across the front office, now we give you the end-to-end view across the enterprise.

>> You brought up a great point, because I was going to actually go there, so this part and what we've been hearing about in the discussions we've already had today was a lot around ecosystem, and a lot about those connectors and bringing... Because to George's point, a lot of people have data all over the place, and Salesforce has legitimately been a target for people to suck data out of, and now with Data Cloud, we heard there's record growth and things of that nature bringing data in. How do you see that ecosystem really growing, I guess you could say, and to George's point, AI working with other AI, for that matter, in that way?

David Schmaier

>> Sure, sure. We have something called the AppExchange, and so we take a platform-centric and ecosystem-centric view to everything we do at Salesforce. In fact, the way I got to Salesforce is I founded and was CEO of a company called Vlocity that built industry apps on the Salesforce platform, and that got acquired by Salesforce in 2020, which is when I became the chief product officer. And Mark and I have a lineage going way back, working at Oracle before Salesforce together, so we've known each other for a long time. And so, I'm a big believer in this platform-centric approach. The AppExchange today has over 6,000 business applications, and what we just announced is a new agent ecosystem, where now you can take our Agentforce platform and you can build your own agents, that's what Workday is doing. So for employee onboarding or employee service, they're building with us on the platform, because they are really the system of record for employee data, we're the system of record for customer data, now we bring both of those together on Agentforce right to the customer, and they can use AI in the flow work to do magical things and have great customer and employee experiences.

>> Should we infer from that that you become the overall agent control framework, and that other application companies are working within essentially the new application model? It used to be application models were stovepiped. Now that you own a multi-agent, multi-vendor system, are you the new application framework?

David Schmaier

>> Well, I would say it this way, potentially. I think we're the new way of bringing agents to the enterprise. So the old way was many of the hyperscalers and many of the AI/LLM companies would say, "Hey ,George, you need to train your own models, and you need to build your own RAG, and you need to get a vector database, and you too can be a computer hobbyist and plug this all in together," and what we're great at doing at Salesforce is making hard things simple. And so, we've done that for 25 years, analytics, we did it with Slack, we did this with the cloud, we did this with mobile, we did this with social, and now we're doing it with agents. Whereas as I said, and like our employee, we call it our advanced force... Because we're literally going to have a whole campground at Dreamforce where you can bring this laptop, if you're a Salesforce customer, and we'll have a couple of experts... It's kind of like the Apple Genius Bar. They'll turn your agent on in one to two hours and they'll show you how to do it, so when you leave there, it'll be in your sandbox. You can go home and test it and look at all the use cases, and then later, if you want go live with it, you can go live with it. So it's very, very cool. So we believe that we've been in this computer hobbyist era of AI, that's why these projects have taken a long time, that's why the returns have not been what everybody expected. And we think this is the new chapter, we think it's going to be easy.

>> You hinted at something, I want to draw you out on this, that hyperscalers were offering their tools, do-it-yourself building blocks, and those were low-level. And we talked about this in the beginning, the applications used to be more difficult to compose. Are we getting to the point where a Salesforce is now the new infrastructure software abstraction layer, the platform-as-a-service and the application, and the cloud is just the hardware infrastructure below that?

David Schmaier

>> Yeah. Let me give you a couple of tangible examples. So the first is Wiley, the textbook company, they're going to be in our keynote. They've been a Salesforce customer for about 10 years. They have a huge spike in demand, our kids all bought their textbooks, we probably bought them from them too years ago.

>> Oh, yeah.

David Schmaier

>> And this is back to school season, so they need all these seasonal workers, but now they don't need them because they're using Agentforce. And so, they found incredible customer satisfaction scores, like 4.4 out of five, and they were able to handle this huge spike in demand, and they were able to do it with Agentforce. Another example is one of my favorite internet franchises, OpenTable, they have 160 million diners. They were going through a chatbot bake-off, and they're a big Salesforce customer, and they said, "Oh, wow, we didn't know about Agentforce, can you tell us more about that?" And they quickly standardized on Agentforce, because chatbots are very brittle, where you had... We have a chatbot framework that we built years ago at Salesforce, and we still have very many successful customers on that. But you basically program, if, then, else, the branching of if the customer says this, then do this, or else do that. Now, with the LLM, you don't have to do any of that, it can reason and it can plan. And so, we're seeing incredible results at Open Table. And then, Disney, which has 17 million visitors a year to the theme parks and 150 million-plus to Disney Plus, I'm a big Disney Plus fan. Now when you go to the theme park, what they're doing through Data Cloud and with agents, is if you are a big Star Wars fan, it can recommend Andor and Obi-Wan is the next series that you should watch on Disney Plus, because it knows who you are, it knows what your preferences are, it knows everything about you, and it can do it in a very human-centric way. And if you want to talk to somebody too, you can, we support humans and agents working together, so it's not either/or, we think it's both. We think that there are some situations where you're going to want to talk to people, like in the theme park, there might be somebody with an iPad guiding you to the next ride that's using Salesforce. But we think the intelligence and the smarts and the AI make that all effortless, it makes it the experience that you want. It's what we call AI that you would really hope to have, an AI that you imagine.

>> So I think what I'm seeing, and again, part of the announcement was 21-plus agents that you guys had announced last week, I'm seeing others that are saying, "Hey, we're going to build all the agents that are in our application stack." And you guys seem to be more, I guess you could say, ecosystem-friendly, in the fact that you're saying, "We're going to build some that make sense and some that we should be having, but we're also going to let others inject their agents, or you can build your own agent into our applications." Is that really, when you look at it with your CPO hat on, is that your roadmap is, hey, I want to enable people to have choice in how the agents are built?

David Schmaier

>> Absolutely. We have certain domain expertise, like our namesake product, our Sales Cloud is the world's number one Salesforce automation product, our Service Cloud is our biggest product in the entire company, it's the number one service product, and Marketing Cloud is the number one marketing product, and Commerce is the number one commerce product, so we think that we know how to build the agents for those disciplines better than other people. But there's no restrictions, you can take our platform, Agent Builder, and other people can build agents too. If they can build a better one, great. But we have, in the AppExchange, 98%-plus of our customers use one or more AppExchange product because it's impossible for us to build everything. A good example is the Workday example. We're not experts at employee onboarding, Workday is. And so, why try to be in the game? They're a great partner, we use their products, they use ours, they're a customer. The same with ADP and payroll. Why not let the experts with the domain expertise build the knowledge and the skills? And if you go to our website, you'll see our Skills Library and what we call our Action Library. So we built hundreds and hundreds of pre-built actions that the agents could take. But in your instance of Agentforce, you can build new agents, and our partners can build new agents too, based on that domain expertise.

>> But just to clarify and make sure it's crystal clear for all the customers who are watching, your identity as a company is changing. You've been an application company for a quarter-century, and you're becoming, it sounds like, a hybrid application and cloud platform company. And so, I imagine that means you're selling not just to functional departmental business leaders, but also the CIO, and you're selling in conjunction, which means you're selling at a much more strategic level. You can go into the business leaders, and you can then also go into the CIO who's responsible for the platform. Is that the new identity?

David Schmaier

>> I would say it's more of an evolution than a pivot, and let me explain why, which is the origins of Salesforce was Sales Cloud, and the very next club was the metadata-driven platform, because we found out that people wanted to configure and customize and add fields or add tabs, we changed things, and then now it's gone way beyond that, with Flow and MuleSoft and other capabilities. So you can literally customize everything if you want to. Part of the magic is we automatically upgrade every single customer, we do three major releases a year, and everybody gets it and it literally just works, there's no big IT upgrade project. So our 12,000 engineers can keep building more and more and more, and customers just get it and they can instantly use it because of this multi-tenant cloud model, so it's really amazing. But we've been a platform company since the very, very beginning. I think what the definition of platform has expanded dramatically. So before, it was about objects, then it was about workflows, then it was about APIs, and Postman would say we have the number one used open APIs in the world, and now it's about agents and AI and the flow work, and we are taking that same platform-centric, ecosystem-centric, view of the world.

>> How do you see that the entire, I guess you could say ecosystem too, a little bit playing off of what George was saying around personas, and AI personas have so traditionally been data science on the traditional ML side of things. Now, with GenAI, you still have the data wrangling and the people who would be enthralled by Data Cloud and things of that nature. How do you see, when you're out with customers, who's in that room now? To George's point, who's in the room when you're out having a conversation about how effective... "Hey, you're going to love this thing we're about to announce next Thursday, it's going to be under NDA, blah, blah, blah. It's Agentforce, and this is what it's going to do, it's going to make agents easy for you." Who in that room is getting excited about that?

David Schmaier

>> So all the customers I named, and many, many more, the biggest companies in the world use Salesforce, and small ones and medium-sized companies too. But we'll typically meet with the CEO, and the CEOs are all saying, "How do I use AI for competitive advantage?" If you come to San Francisco, Uber was a revelation and they're a huge customer and a great example of how you disrupt an industry, and they run their whole company on Salesforce. Amazon's another great company and they're one of our biggest customers in the world, and they disrupted their industry and they run their whole company based on Salesforce. But you might see all these Waymo cars now and say there's a new way to get around, and it's through these autonomous vehicles. I took one yesterday and it was a great experience, and it just shows you that that's going to happen in every single industry. So that's what Agentforce is about, is bringing that level of disruption. Now, to your question of we have that meeting and the CEO says, "Yeah, we need to be the Waymo of our industry, or the Uber of our industry, let's do it." And then, he turns to his CIO or CTO and says, "Make it happen." And that's where we start spending a lot of our time. And there, I think, is where there's been a little bit of a, I would say, early attempts at this with DIY AI, so do-it-yourself AI, where people are trying to cobble this together and build it themselves. But it's pretty hard to... We have a PhD from Stanford in machine learning and AI who runs our AI research lab, and we have a team of PhDs who work for him, it's pretty hard to fill the team like that. Some companies have that, but most don't. And what we would say now with Agentforce and our low-code platform is you don't need that, you just need to use it, you don't need to build it.

>> Do you see the rapidly-evolving capabilities of the underlying LLMs almost like the rapidly evolving speeds of the servers or the capabilities of the relational databases several decades ago? The current Agentforce models, I think, are built on an open source MISRA or whatever, and we've seen, just in the past few months, dramatic changes in capabilities for planning for tool use. How might that show up in the ambition of the agents that you can offer to the customers?

David Schmaier

>> Yeah, we're in a brand new world. In fact, they don't run on MISRA, we have what's called an open LLM framework. So actually, you can pick the LLM of your choice, and we have certain ones certified, OpenAI, Anthropic, Amazon's LLMs, and others. And so, we let customers have a choice there, just like we let them have a choice with the ecosystem approach. But what we're finding, and Agentforce is a perfect example, is we're beating the results people were seeing with the custom application of LLMs. And the reason we're beating it is because we understand all the data, we understand all the metadata, which is the real semantic meaning of what's really going on, therefore we understand the context of every single customer interaction across every channel. So it's really not the LLM that's the magic, it's putting it all together with Data Cloud, and we've also introduced with Agentforce something we call our Atlas Planning and Reasoning Engine, this is really cool. Our AI research team run by this Stanford professor built this, and we built this in partnership with some of our early customers, a big bank, one of the leading healthcare companies in America, ADP, Disney is a company we worked with on this. And this really looks at the meaning of what people want to do, and then plans the outcome of what the actions are, according to the action library, according to this customer profile. So you don't need to build all this stuff anymore, that was the chapter one of a 20 chapter book on GenAI, now you can just use low-code and no-code. And we pre-built these agents to do a whole variety of things. You can extend them yourself without programming them, you can add new things. We're actually going to show that in the keynote tomorrow, where we show an agent that solves 90% of a customer's problem but couldn't do one thing, and we go in and with a couple of clicks and we add one more action, a custom action, and then it can do 100% of what the customer wants and for every subsequent customer. So that's how easy this should be. Well, guess what? That's how easy it is today with Agentforce.

>> One of the things that we're learning when we talk to the players in this, both at the application level, the tools vendors, the model vendors, is that in the era of symbolic code, which was the last 60 years, we only could reach the high-end of the power curve, the 20% of processes that you could specify. But now, the promise is we can reach the 80% that you could only learn. And so, how does that change your ambitions for Salesforce as an application platform in terms of what you can reach? But then, are you starting to think about advising customers not just on being more ambitious in what they build, but how they run their companies, how they transform their companies?

David Schmaier

>> Sure. Our view is that the LLMs are really important, and we work with all the LLM companies, and many of them run their whole companies on Slack and Salesforce. Actually, I think all of them do. And so, they're all customers and partners of ours, and we'll plug in more of them into our LLM gateways, because we want people to be able to... We have to certify them, and we have a trust boundary to make sure that everything that happens within our trust boundary is according to ethical use and it screens out bias. So we take that part of our trust core value really seriously. But we do think LLMs over time are going to become more of a commodity, and if you go to Hugging Face and you look at the leaderboard, they're all hovering around in the mid-90% accuracies, they're going to get to the high 90s. And then, at some point, the differences are going to be imperceptible to the naked eye, 98.7% accuracy versus 98.5%. This is like the old database wars with who had this many transactions per second, they sort of became all the same, and that's our view. The real magic again is putting the data together, the real magic is the use cases. The other thing we're finding incredible interest in is all the advanced RAG techniques. So we're doing something called Ensemble RAG, which is beyond basic RAG, that really gets us far higher accuracy. And then, the final thing that we've done with our Data Cloud is we think what our customers want is to take structured data and unstructured data and put that all together in the output, and we think that that's also a unique advantage that we have. We have all the conversations in Slack, we have threaded discussions within Salesforce, and we have all the structured data in Data Cloud and in all of our transactional apps.

>> So one last thing, because I think I was asked about this prior to coming here, and it's not necessarily on the AI topic, but it really ties into the ecosystem and into protecting your data and stuff like that, as you're going out and you've made an intent to acquire Own Company, which used to be OwnBackup and things of that nature, how does that fit in with the strategy going... How do you see, assuming... You don't have to assume anything because it's not done yet, I know that. But where does that fit into your strategy going forward of protecting the data and making data core to what you're doing? We all know AI runs on data, so that would seem like a very key piece for the platform.

David Schmaier

>> Yeah. You just gave the answer, which is AI runs on data, and so therefore, the data is critical. We now have rotated to Agentforce ads, but we were running ads with Matthew McConaughey saying, "Data is the new gold." We really believe that, that's the gold of the AI age, and it needs to be mined and it needs to be harmonized and unified. And once you have that unified view of the world, sometimes black swan things happen and you need to back up and archive that data. And so, we provided some level of solutions, but just like I said before, we have an open platform, where we have ISVs like Own, and it was originally called OwnBackup, they have five products, actually, the backup is their biggest product, but they do archiving, they do data security and data masking, which will augment our shield products. And once the transaction's approved, it's not approved yet, not finalized yet, but once it's done, we think it's just one more way to reinforce... We have five core values in the company, and the number one value is trust, and so people trust putting their customer data in the cloud with Salesforce, and now that trust will just be that much further, because now you can back it up and you can archive it and you can screen out certain selective content with some of their other products. And so, we think it makes perfect sense.

>> Let me just see if I can play back something you said just before about where you can add value and where you can capture value. The Frontier foundation model vendors face brutal economics in the cost of compute, the cost of the researchers, the cost of computer training, and even for serving.

David Schmaier

>> You mean you don't want to buy billions or tens of billions of GPUs?

>> Well, Jensen is on tomorrow, so I don't know that we want to...

David Schmaier

>> No, he's a great partner and a great customer-

>> Yeah....

David Schmaier

>> so power to him. He's got a great business model.

>> Yes, he does.

>> He makes the money. But the Frontier model vendors face, for now, brutal economics also because for whatever they build, their expertise can be distilled into smaller models, even the OpenAI stuff, because Microsoft gets to do that, so Llama, Microsoft, others. If these Frontier models, let's say, are trained on a petabyte of data, or something like that, then they can generate maybe math and code. But the expertise for any particular function or industry is captured, let's say, in J.P. Morgan's 250 petabytes of data. So there's two sources to capture value, the tools to organize that expertise and to fine-tune the models into agents that capture and embody that expertise, or the customer who then turns that raw data into something repeatable. And I want to get your reaction too, is that how you see the market going forward in terms of where the value gets captured?

David Schmaier

>> Sure. We have 15 industry clouds, so digital transformation is an industry-specific sport. My last company built six of the 15 at Vlocity, so this is an area I've spent some time in. AI in the flow work is really important, AI in the flow of industry work is a little nuanced, so we add hundreds of data objects to these industry clouds. Maybe you care about the account balance or the credit score of someone in banking, or maybe you care about the billing profile and how many residents somebody has for the energy or the telecom or media business, so you turn on the service in the right location at the right time. And so, each one of these industries has trade promotions data and consumer goods companies, so each one has different data. We can use that data to inform the AI in the flow work, because we've codified the best practices for sales, marketing, service, in all 15 of these industries. So we've already got all of that there to light up with generative AI. Now, we do think the foundation model, and I was kidding, they're partners of ours, they're spending all the money with Jensen so we can just use that stuff, which is great and we just take advantage of that. But we also have the capability in Salesforce research to build our own models. So we've built a CodeGen product for our Apex programming language, which is a JavaScript variant, we've built a number of models, and we see a lot of action in the future in small models and domain-specific models, and so we are working on some new ones in that area that we think are going to be very exciting.

>> I think that is a perfect place to leave it, because I think we violently agree with you on SLMs, or small language models, and how they're going to really impact going forward, and we'll have to continue this at another time.

David Schmaier

>> Part two.

>> Yes. But thanks, David, for coming on board, this has been highly enlightening, really appreciate it, and really looking forward to this week here.

David Schmaier

>> No, loved the conversation, thank you for having me, and let's have a great Dreamforce.

>> Absolutely. And why don't you have a great Dreamforce and join us again tomorrow for more Dreamforce 2024, live from the NYSE, our CUBE on location, just north of Moscone. See you tomorrow. Enjoy. Bye.

David Schmaier | Salesforce

search

>> Hello, welcome back to Dreamforce 2024, live from the NYSE, where we have set up just outside of Moscone to really bring to you what's going on and what you should expect. This week we're really excited, because, again, we get some of the best guests that Salesforce has been really great about bringing on here. And we have David Schmaier, who is the president and chief product officer with Salesforce, who's going to help us break it down. I'm also joined by George Gilbert, who's our data guru-in-chief here, and we're really going to dig in on some of the announcements that's been going on. I think one of the big things that was talked about last week was really Agentforce and the announcements around that. Help us understand where you're thinking and how to get... People always talk about bots and getting past bots, so how does this announcement really help with that?

David Schmaier

>> Sure. And by the way, thank you for having me here today, really appreciate it. Great to see you both. If there's three things that you take away from Dreamforce this year, it's Agentforce, Agentforce, and Agentforce. So we're going all in on humans and AI agents working together to help customers, and we really think this is a big story because a lot of companies have been trying to build their own AI, and we don't think this is something you want to do it yourself. If you go to a surgeon, you don't do this yourself, you go to somebody who really knows what they're doing. And what we've done is built this out-of-the-box so you can literally turn our agents on in hours, and not take months or even years to build the AI yourself.

>> So let me just ask at the top-level this value proposition, there's always this tension between an application that is out-of-the-box value that's packaged and the desire to allow it to be building blocks for custom applications, and no one's ever really been able to make that both a dessert topping and a floor wax, to borrow a phrase. Agents seem to get us closer to that. What is it about agents that allow you to provide the best of both, and what did you have to do to the foundational components of Salesforce to enable that?

David Schmaier

>> That's a great question, George. We really have built Agentforce on top of the foundation we built over 25 years, so multi-tenant, in the cloud, proven, scalable, trusted. On top of that, we built something new in the last several years called our Data Cloud, which allows you to basically integrate and harmonize and unify all the data together. And our view is the data powers the AI and the agents, and the agents power the customer experience for the next generation. So we've really brought it all together with Agentforce, and in my opinion, I've been working on what we call CRM software for over three decades, this is the dream of CRM, where you and I, the three of us all work together and we put data into the system, and the system magically does things for us. Now, the AI agents can do that magic.

>> It would seem that normalization of that data is a big task, and again, having been and used Salesforce in multiple different ways in multiple different parts of the cloud, Salesforce Cloud, how did you really achieve that helps people normalize that data so that AI can really be actionable, really be trusted, within that?

David Schmaier

>> Sure. We've been working on AI for over a decade at Salesforce. We have a Stanford professor of AI and machine learning who runs our research center. So we've been working on this for literal years, and we do over two trillion AI transactions every week. But when GenAI came out and ChatGPT came out, the whole industry changed. And so, what we've found is this is a real opportunity to double down and go all in on agents and AI in the flow work, so that's what we've done, and what I'm most excited about are the customers. So customers like OpenTable and Disney and ADP and Wiley and Saks Fifth Avenue, you're going to hear from them this week, you're going to hear their stories, they're going to talk about how the agents have literally transformed and revolutionized the way they interact with their customers, so they can do what has really been our vision for 25 years, that we help companies connect with their customers in entirely new ways, and Agentforce is definitely an entirely new way.

>> So Data Cloud, it took a few years to pull together a historical system of truth, and even some harmonized integration to the real-time system of truth in the operational applications. But what seems to tie it all together is you have a common language for Customer 360 and the customer journey, that was critical, and none of the other data clouds really have that. But then, you also have the common language for the business processes, which now become the tools or actions that the agents can invoke to get work done. Now, as far as we understand, no one else has all those ingredients, because you needed those harmonized actions not just to get stuff done, but by connecting it to the applications, they can learn from the outcomes, and so the agents get smarter over time.

David Schmaier

>> That's exactly right. I think that's very, very well said. So we have always had the full Customer 360 view, and some of the other CRM players give you what what we call Customer 360, they give you a customer 90 or Customer 180, only Salesforce gives you the end-to-end view across every single channel of communication, across sales, marketing, service, and commerce. But now, the data lake, the Data Cloud, takes it further, because we can bring in SAP data, we can bring in billing data. If somebody clicks on a certain product page on your website, you instantly get that. That's a signal that we can send to the right salespeople and say, "Hey, somebody just clicked on this page, they might want to buy this particular product." That would be good to know instantaneously, because leads are sometimes perishable. So we've connected all this together. The amazing thing about Agentforce is we have over 200,000 customers at Salesforce, and it's all built on these core fundamental technologies, like at the data layer and at the workflow layer, our flow automation layer does 44 billion workflows a day, and so it's used and beloved by all of our customers. So if you use Agentforce, now with a little bit of upskilling... If you know how to use Salesforce, pardon me, now you know how to use Agentforce. And so, we did a training yesterday of our own people who we're going to turn AI agents on for people right at Dreamforce, and we had our own people turning on agents in one to two hours, working, really smart, intelligent, conversational agents in one to two hours.

>> You said something interesting, where you make it so easy to harmonize all the customer data in one language, Customer 360, the journey, and that might make it easy, there's almost like a network effect because it draws in non-customer data that's related to the customer. So is that one way of bringing, turning this, what was a customer system of truth, into a broader system of truth, and then through connectors, interacting with other operational applications? In other words, is the aspiration to become the application platform for agents beyond just the customer surface points?

David Schmaier

>> Sure. Well, the data lake, our Data Cloud, allows you to... We now have over 400 connectors, and with MuleSoft, which is the number one API management system, we can get you to thousands and thousands of applications. So we can virtually connect to anything, but we can also connect to other data lakes, like Snowflake and Databricks and Redshift and BigQuery and Azure Fabric data, so we can connect to any of these other data lakes. And we just announced a big partnership with Workday and with IBM and other big partners that are also joining what we call the Zero Copy Alliance, where you don't need to copy all the data into our Data Cloud, we can just point to it, but give you a logical view of that. So what you said, George, is exactly right, that what we used to give you, the end-to-end view across the front office, now we give you the end-to-end view across the enterprise.

>> You brought up a great point, because I was going to actually go there, so this part and what we've been hearing about in the discussions we've already had today was a lot around ecosystem, and a lot about those connectors and bringing... Because to George's point, a lot of people have data all over the place, and Salesforce has legitimately been a target for people to suck data out of, and now with Data Cloud, we heard there's record growth and things of that nature bringing data in. How do you see that ecosystem really growing, I guess you could say, and to George's point, AI working with other AI, for that matter, in that way?

David Schmaier

>> Sure, sure. We have something called the AppExchange, and so we take a platform-centric and ecosystem-centric view to everything we do at Salesforce. In fact, the way I got to Salesforce is I founded and was CEO of a company called Vlocity that built industry apps on the Salesforce platform, and that got acquired by Salesforce in 2020, which is when I became the chief product officer. And Mark and I have a lineage going way back, working at Oracle before Salesforce together, so we've known each other for a long time. And so, I'm a big believer in this platform-centric approach. The AppExchange today has over 6,000 business applications, and what we just announced is a new agent ecosystem, where now you can take our Agentforce platform and you can build your own agents, that's what Workday is doing. So for employee onboarding or employee service, they're building with us on the platform, because they are really the system of record for employee data, we're the system of record for customer data, now we bring both of those together on Agentforce right to the customer, and they can use AI in the flow work to do magical things and have great customer and employee experiences.

>> Should we infer from that that you become the overall agent control framework, and that other application companies are working within essentially the new application model? It used to be application models were stovepiped. Now that you own a multi-agent, multi-vendor system, are you the new application framework?

David Schmaier

>> Well, I would say it this way, potentially. I think we're the new way of bringing agents to the enterprise. So the old way was many of the hyperscalers and many of the AI/LLM companies would say, "Hey ,George, you need to train your own models, and you need to build your own RAG, and you need to get a vector database, and you too can be a computer hobbyist and plug this all in together," and what we're great at doing at Salesforce is making hard things simple. And so, we've done that for 25 years, analytics, we did it with Slack, we did this with the cloud, we did this with mobile, we did this with social, and now we're doing it with agents. Whereas as I said, and like our employee, we call it our advanced force... Because we're literally going to have a whole campground at Dreamforce where you can bring this laptop, if you're a Salesforce customer, and we'll have a couple of experts... It's kind of like the Apple Genius Bar. They'll turn your agent on in one to two hours and they'll show you how to do it, so when you leave there, it'll be in your sandbox. You can go home and test it and look at all the use cases, and then later, if you want go live with it, you can go live with it. So it's very, very cool. So we believe that we've been in this computer hobbyist era of AI, that's why these projects have taken a long time, that's why the returns have not been what everybody expected. And we think this is the new chapter, we think it's going to be easy.

>> You hinted at something, I want to draw you out on this, that hyperscalers were offering their tools, do-it-yourself building blocks, and those were low-level. And we talked about this in the beginning, the applications used to be more difficult to compose. Are we getting to the point where a Salesforce is now the new infrastructure software abstraction layer, the platform-as-a-service and the application, and the cloud is just the hardware infrastructure below that?

David Schmaier

>> Yeah. Let me give you a couple of tangible examples. So the first is Wiley, the textbook company, they're going to be in our keynote. They've been a Salesforce customer for about 10 years. They have a huge spike in demand, our kids all bought their textbooks, we probably bought them from them too years ago.

>> Oh, yeah.

David Schmaier

>> And this is back to school season, so they need all these seasonal workers, but now they don't need them because they're using Agentforce. And so, they found incredible customer satisfaction scores, like 4.4 out of five, and they were able to handle this huge spike in demand, and they were able to do it with Agentforce. Another example is one of my favorite internet franchises, OpenTable, they have 160 million diners. They were going through a chatbot bake-off, and they're a big Salesforce customer, and they said, "Oh, wow, we didn't know about Agentforce, can you tell us more about that?" And they quickly standardized on Agentforce, because chatbots are very brittle, where you had... We have a chatbot framework that we built years ago at Salesforce, and we still have very many successful customers on that. But you basically program, if, then, else, the branching of if the customer says this, then do this, or else do that. Now, with the LLM, you don't have to do any of that, it can reason and it can plan. And so, we're seeing incredible results at Open Table. And then, Disney, which has 17 million visitors a year to the theme parks and 150 million-plus to Disney Plus, I'm a big Disney Plus fan. Now when you go to the theme park, what they're doing through Data Cloud and with agents, is if you are a big Star Wars fan, it can recommend Andor and Obi-Wan is the next series that you should watch on Disney Plus, because it knows who you are, it knows what your preferences are, it knows everything about you, and it can do it in a very human-centric way. And if you want to talk to somebody too, you can, we support humans and agents working together, so it's not either/or, we think it's both. We think that there are some situations where you're going to want to talk to people, like in the theme park, there might be somebody with an iPad guiding you to the next ride that's using Salesforce. But we think the intelligence and the smarts and the AI make that all effortless, it makes it the experience that you want. It's what we call AI that you would really hope to have, an AI that you imagine.

>> So I think what I'm seeing, and again, part of the announcement was 21-plus agents that you guys had announced last week, I'm seeing others that are saying, "Hey, we're going to build all the agents that are in our application stack." And you guys seem to be more, I guess you could say, ecosystem-friendly, in the fact that you're saying, "We're going to build some that make sense and some that we should be having, but we're also going to let others inject their agents, or you can build your own agent into our applications." Is that really, when you look at it with your CPO hat on, is that your roadmap is, hey, I want to enable people to have choice in how the agents are built?

David Schmaier

>> Absolutely. We have certain domain expertise, like our namesake product, our Sales Cloud is the world's number one Salesforce automation product, our Service Cloud is our biggest product in the entire company, it's the number one service product, and Marketing Cloud is the number one marketing product, and Commerce is the number one commerce product, so we think that we know how to build the agents for those disciplines better than other people. But there's no restrictions, you can take our platform, Agent Builder, and other people can build agents too. If they can build a better one, great. But we have, in the AppExchange, 98%-plus of our customers use one or more AppExchange product because it's impossible for us to build everything. A good example is the Workday example. We're not experts at employee onboarding, Workday is. And so, why try to be in the game? They're a great partner, we use their products, they use ours, they're a customer. The same with ADP and payroll. Why not let the experts with the domain expertise build the knowledge and the skills? And if you go to our website, you'll see our Skills Library and what we call our Action Library. So we built hundreds and hundreds of pre-built actions that the agents could take. But in your instance of Agentforce, you can build new agents, and our partners can build new agents too, based on that domain expertise.

>> But just to clarify and make sure it's crystal clear for all the customers who are watching, your identity as a company is changing. You've been an application company for a quarter-century, and you're becoming, it sounds like, a hybrid application and cloud platform company. And so, I imagine that means you're selling not just to functional departmental business leaders, but also the CIO, and you're selling in conjunction, which means you're selling at a much more strategic level. You can go into the business leaders, and you can then also go into the CIO who's responsible for the platform. Is that the new identity?

David Schmaier

>> I would say it's more of an evolution than a pivot, and let me explain why, which is the origins of Salesforce was Sales Cloud, and the very next club was the metadata-driven platform, because we found out that people wanted to configure and customize and add fields or add tabs, we changed things, and then now it's gone way beyond that, with Flow and MuleSoft and other capabilities. So you can literally customize everything if you want to. Part of the magic is we automatically upgrade every single customer, we do three major releases a year, and everybody gets it and it literally just works, there's no big IT upgrade project. So our 12,000 engineers can keep building more and more and more, and customers just get it and they can instantly use it because of this multi-tenant cloud model, so it's really amazing. But we've been a platform company since the very, very beginning. I think what the definition of platform has expanded dramatically. So before, it was about objects, then it was about workflows, then it was about APIs, and Postman would say we have the number one used open APIs in the world, and now it's about agents and AI and the flow work, and we are taking that same platform-centric, ecosystem-centric, view of the world.

>> How do you see that the entire, I guess you could say ecosystem too, a little bit playing off of what George was saying around personas, and AI personas have so traditionally been data science on the traditional ML side of things. Now, with GenAI, you still have the data wrangling and the people who would be enthralled by Data Cloud and things of that nature. How do you see, when you're out with customers, who's in that room now? To George's point, who's in the room when you're out having a conversation about how effective... "Hey, you're going to love this thing we're about to announce next Thursday, it's going to be under NDA, blah, blah, blah. It's Agentforce, and this is what it's going to do, it's going to make agents easy for you." Who in that room is getting excited about that?

David Schmaier

>> So all the customers I named, and many, many more, the biggest companies in the world use Salesforce, and small ones and medium-sized companies too. But we'll typically meet with the CEO, and the CEOs are all saying, "How do I use AI for competitive advantage?" If you come to San Francisco, Uber was a revelation and they're a huge customer and a great example of how you disrupt an industry, and they run their whole company on Salesforce. Amazon's another great company and they're one of our biggest customers in the world, and they disrupted their industry and they run their whole company based on Salesforce. But you might see all these Waymo cars now and say there's a new way to get around, and it's through these autonomous vehicles. I took one yesterday and it was a great experience, and it just shows you that that's going to happen in every single industry. So that's what Agentforce is about, is bringing that level of disruption. Now, to your question of we have that meeting and the CEO says, "Yeah, we need to be the Waymo of our industry, or the Uber of our industry, let's do it." And then, he turns to his CIO or CTO and says, "Make it happen." And that's where we start spending a lot of our time. And there, I think, is where there's been a little bit of a, I would say, early attempts at this with DIY AI, so do-it-yourself AI, where people are trying to cobble this together and build it themselves. But it's pretty hard to... We have a PhD from Stanford in machine learning and AI who runs our AI research lab, and we have a team of PhDs who work for him, it's pretty hard to fill the team like that. Some companies have that, but most don't. And what we would say now with Agentforce and our low-code platform is you don't need that, you just need to use it, you don't need to build it.

>> Do you see the rapidly-evolving capabilities of the underlying LLMs almost like the rapidly evolving speeds of the servers or the capabilities of the relational databases several decades ago? The current Agentforce models, I think, are built on an open source MISRA or whatever, and we've seen, just in the past few months, dramatic changes in capabilities for planning for tool use. How might that show up in the ambition of the agents that you can offer to the customers?

David Schmaier

>> Yeah, we're in a brand new world. In fact, they don't run on MISRA, we have what's called an open LLM framework. So actually, you can pick the LLM of your choice, and we have certain ones certified, OpenAI, Anthropic, Amazon's LLMs, and others. And so, we let customers have a choice there, just like we let them have a choice with the ecosystem approach. But what we're finding, and Agentforce is a perfect example, is we're beating the results people were seeing with the custom application of LLMs. And the reason we're beating it is because we understand all the data, we understand all the metadata, which is the real semantic meaning of what's really going on, therefore we understand the context of every single customer interaction across every channel. So it's really not the LLM that's the magic, it's putting it all together with Data Cloud, and we've also introduced with Agentforce something we call our Atlas Planning and Reasoning Engine, this is really cool. Our AI research team run by this Stanford professor built this, and we built this in partnership with some of our early customers, a big bank, one of the leading healthcare companies in America, ADP, Disney is a company we worked with on this. And this really looks at the meaning of what people want to do, and then plans the outcome of what the actions are, according to the action library, according to this customer profile. So you don't need to build all this stuff anymore, that was the chapter one of a 20 chapter book on GenAI, now you can just use low-code and no-code. And we pre-built these agents to do a whole variety of things. You can extend them yourself without programming them, you can add new things. We're actually going to show that in the keynote tomorrow, where we show an agent that solves 90% of a customer's problem but couldn't do one thing, and we go in and with a couple of clicks and we add one more action, a custom action, and then it can do 100% of what the customer wants and for every subsequent customer. So that's how easy this should be. Well, guess what? That's how easy it is today with Agentforce.

>> One of the things that we're learning when we talk to the players in this, both at the application level, the tools vendors, the model vendors, is that in the era of symbolic code, which was the last 60 years, we only could reach the high-end of the power curve, the 20% of processes that you could specify. But now, the promise is we can reach the 80% that you could only learn. And so, how does that change your ambitions for Salesforce as an application platform in terms of what you can reach? But then, are you starting to think about advising customers not just on being more ambitious in what they build, but how they run their companies, how they transform their companies?

David Schmaier

>> Sure. Our view is that the LLMs are really important, and we work with all the LLM companies, and many of them run their whole companies on Slack and Salesforce. Actually, I think all of them do. And so, they're all customers and partners of ours, and we'll plug in more of them into our LLM gateways, because we want people to be able to... We have to certify them, and we have a trust boundary to make sure that everything that happens within our trust boundary is according to ethical use and it screens out bias. So we take that part of our trust core value really seriously. But we do think LLMs over time are going to become more of a commodity, and if you go to Hugging Face and you look at the leaderboard, they're all hovering around in the mid-90% accuracies, they're going to get to the high 90s. And then, at some point, the differences are going to be imperceptible to the naked eye, 98.7% accuracy versus 98.5%. This is like the old database wars with who had this many transactions per second, they sort of became all the same, and that's our view. The real magic again is putting the data together, the real magic is the use cases. The other thing we're finding incredible interest in is all the advanced RAG techniques. So we're doing something called Ensemble RAG, which is beyond basic RAG, that really gets us far higher accuracy. And then, the final thing that we've done with our Data Cloud is we think what our customers want is to take structured data and unstructured data and put that all together in the output, and we think that that's also a unique advantage that we have. We have all the conversations in Slack, we have threaded discussions within Salesforce, and we have all the structured data in Data Cloud and in all of our transactional apps.

>> So one last thing, because I think I was asked about this prior to coming here, and it's not necessarily on the AI topic, but it really ties into the ecosystem and into protecting your data and stuff like that, as you're going out and you've made an intent to acquire Own Company, which used to be OwnBackup and things of that nature, how does that fit in with the strategy going... How do you see, assuming... You don't have to assume anything because it's not done yet, I know that. But where does that fit into your strategy going forward of protecting the data and making data core to what you're doing? We all know AI runs on data, so that would seem like a very key piece for the platform.

David Schmaier

>> Yeah. You just gave the answer, which is AI runs on data, and so therefore, the data is critical. We now have rotated to Agentforce ads, but we were running ads with Matthew McConaughey saying, "Data is the new gold." We really believe that, that's the gold of the AI age, and it needs to be mined and it needs to be harmonized and unified. And once you have that unified view of the world, sometimes black swan things happen and you need to back up and archive that data. And so, we provided some level of solutions, but just like I said before, we have an open platform, where we have ISVs like Own, and it was originally called OwnBackup, they have five products, actually, the backup is their biggest product, but they do archiving, they do data security and data masking, which will augment our shield products. And once the transaction's approved, it's not approved yet, not finalized yet, but once it's done, we think it's just one more way to reinforce... We have five core values in the company, and the number one value is trust, and so people trust putting their customer data in the cloud with Salesforce, and now that trust will just be that much further, because now you can back it up and you can archive it and you can screen out certain selective content with some of their other products. And so, we think it makes perfect sense.

>> Let me just see if I can play back something you said just before about where you can add value and where you can capture value. The Frontier foundation model vendors face brutal economics in the cost of compute, the cost of the researchers, the cost of computer training, and even for serving.

David Schmaier

>> You mean you don't want to buy billions or tens of billions of GPUs?

>> Well, Jensen is on tomorrow, so I don't know that we want to...

David Schmaier

>> No, he's a great partner and a great customer-

>> Yeah....

David Schmaier

>> so power to him. He's got a great business model.

>> Yes, he does.

>> He makes the money. But the Frontier model vendors face, for now, brutal economics also because for whatever they build, their expertise can be distilled into smaller models, even the OpenAI stuff, because Microsoft gets to do that, so Llama, Microsoft, others. If these Frontier models, let's say, are trained on a petabyte of data, or something like that, then they can generate maybe math and code. But the expertise for any particular function or industry is captured, let's say, in J.P. Morgan's 250 petabytes of data. So there's two sources to capture value, the tools to organize that expertise and to fine-tune the models into agents that capture and embody that expertise, or the customer who then turns that raw data into something repeatable. And I want to get your reaction too, is that how you see the market going forward in terms of where the value gets captured?

David Schmaier

>> Sure. We have 15 industry clouds, so digital transformation is an industry-specific sport. My last company built six of the 15 at Vlocity, so this is an area I've spent some time in. AI in the flow work is really important, AI in the flow of industry work is a little nuanced, so we add hundreds of data objects to these industry clouds. Maybe you care about the account balance or the credit score of someone in banking, or maybe you care about the billing profile and how many residents somebody has for the energy or the telecom or media business, so you turn on the service in the right location at the right time. And so, each one of these industries has trade promotions data and consumer goods companies, so each one has different data. We can use that data to inform the AI in the flow work, because we've codified the best practices for sales, marketing, service, in all 15 of these industries. So we've already got all of that there to light up with generative AI. Now, we do think the foundation model, and I was kidding, they're partners of ours, they're spending all the money with Jensen so we can just use that stuff, which is great and we just take advantage of that. But we also have the capability in Salesforce research to build our own models. So we've built a CodeGen product for our Apex programming language, which is a JavaScript variant, we've built a number of models, and we see a lot of action in the future in small models and domain-specific models, and so we are working on some new ones in that area that we think are going to be very exciting.

>> I think that is a perfect place to leave it, because I think we violently agree with you on SLMs, or small language models, and how they're going to really impact going forward, and we'll have to continue this at another time.

David Schmaier

>> Part two.

>> Yes. But thanks, David, for coming on board, this has been highly enlightening, really appreciate it, and really looking forward to this week here.

David Schmaier

>> No, loved the conversation, thank you for having me, and let's have a great Dreamforce.

>> Absolutely. And why don't you have a great Dreamforce and join us again tomorrow for more Dreamforce 2024, live from the NYSE, our CUBE on location, just north of Moscone. See you tomorrow. Enjoy. Bye.