Dreamforce 2025 | Muddu Sudhakar, Salesforce

Clips
More from Dreamforce 2025

Muddu Sudhakar

SVP & GM, IT & HR Service

Salesforce

play_circle_outline Introduction of Moodhu Siddhara and his new role at Salesforce.

play_circle_outline Transforming IT and HR Services: Moodhu Siddhara's Vision for Enhanced Customer Experience with AI Agents at Salesforce

play_circle_outline Evolution of IT service management through automation and AI integration.

play_circle_outline Differentiating Salesforce through its data cloud and AgentForce platform.

Info
Transcript

Muddu Sudhakar, Salesforce

Muddu Sudhakar

SVP & GM, IT & HR Service Salesforce

At Dreamforce 2025, Muddu Sudhakar, senior vice president and general manager of IT & HR service at Salesforce, joins theCUBE Research’s John Furrier and George Gilbert to unpack how Salesforce is operationalizing the agentic enterprise. Sudhakar details Agentforce IT Service Desk, a Slack-first, Teams-ready experience layered on Salesforce Data Cloud’s zero-copy foundation and contextual modeling. He explains why subject-matter agents (for software access, networking, root-cause analysis and more) paired with flows, actions and tools mark the shift from tick... Read more

explore Keep Exploring

What is Moodhu Siddhara's current role and what are his goals and focus areas at Salesforce? add

What are the speaker's thoughts on their new role and the opportunities it presents in IT service management? add

What is your vision for the evolution of IT service management in the context of your new role as SVP at Salesforce? add

What are the advantages of your system compared to the competition? add

bolt Powered by CUBE AI

Muddu Sudhakar, Salesforce

search

>> Welcome back, everyone. I'm John Furrier, the host of theCUBE with George Gilbert, the industry covering Salesforce and data for theCUBE research. We are here streaming live from Dreamforce 2025 where the latest product innovations from Salesforce and now AgentForce is the center of the action. Agentforce, fill in the blank, they got it all and in productivity, it's a tech platform and I'm excited for this next guest, CUBE alumni, now senior leader at Salesforce, Moodhu Siddhara, senior vice president, general manager, IT and HR service at Salesforce. Moodhu, great to see you, friend of theCUBE. You've been probably on a dozen times on theCUBE. Now you're a senior leader at Salesforce. Congratulations. So first of all, first question for you is love your new roles, right in your wheelhouse. What are you working on? What's your focus? What are you optimizing for? What are some of your goals?

Muddu Sudhakar

>> First, I want to thank you, John. You and me go back to 15 years, so thank you for having me. It's been a great experience working with theCUBE. And again, great opportunity at Salesforce. Thanks to Mark Benioff, Jujar Singh and Kishan Chetan. I'm really excited to be about this opportunity. And the reason I took over this role is very clear. Look, this will be my fourth time doing ITSM and I don't want to call the word it, the word ITSM. The new world is AgentForce IT service, right? So I think for me, the way which I see the role is how do we make customers happy in terms of experience? How do you use AI agents on top of AgentForce to automate as many conversations? I want to launch a proactive service, which is conversational first, agent first, and meet the user where they are. That is Slack first and Teams ready at the launch and that's the opportunity, I like it.

>> We have talked on theCUBE if you go back in the archive on ourcubeai.com now publicly available, we've talked about how agents in AI early on how AI and the role of data would change the ITSM, the IT service management business. And because of the data, the richness of the data, the quality of the data, and many times, so now it's here. Okay, so take us to the current state of where we are and where you are right now with Salesforce because the conversations we've been having is harmonizing, unifying data, structured/unstructured. It's a gold mine of opportunity to one, create value and extract value. What's your vision today, now that you're in the SVP role at Salesforce of how you see the IT service management, if we're going to use that word, certainly an existing data market... how do you see that evolving?

Muddu Sudhakar

>> Very good question. Look, the way which I see the world is in multiple areas. The IT service itself that we see, today's first thing is we have to do is make sure that most mundane tasks are automated, right? Employees, when they have a request 24 by 7, you have to have AI agents, but they need data as you said. Without data, nothing will happen. And on top of the data, you need the context. What Salesforce has done with the Data Cloud, Data360 and Agent360, so marvelous. Before, what happened is data is in a silo. Knowledge articles are in legacy IT system systems. They're not outside. Sometimes it could be in some other third party systems, and that is where the zero-copy data cloud really gave the foundation. Then the context layer came in. Then the agent post platform came to build the agents. I call them subject matter agents. So I have an agent for software collaboration, application provisioning, software access, network agent, root cause analysis agent. So you are building 25, 30 agents on top of the platform to deliver and with the workflows. See, what really changed even after number 2022, after ChatGPTs. this year is all about agents and agenting. That means requires workflows, actions and tools. That's what we are delivering today in the GA. But where we want to go from here? We want to go into IT Asset Management, that's hardware, software. We want to go into IT compliance, GRC, we want to go into future world of enterprise service management, and finally go into AI observability, AI ops. This is a large investment. Marc Benioff, Kishan and Jujar made an investment. This is not one-time investment smart move. This is like thousands of engineer-

>> Smart move. It's an obvious move.

Muddu Sudhakar

>> Yes.

George Gilbert

>> Maybe let me ask, I want to drill down on a couple of things. You mentioned like the subject matter agents sounded really interesting in the observability. We talked to Madhav and others on the agent force team, that first big data was about click streams coming from systems of engagement and trying to understand the customer experience, and now we need the agent exhaust integrated with the existing data cloud. Tell us a little bit more about how IT is capturing that exhaust, but making it part of a new form of analytics where there's a learning loop for the agents. So it's partly operations, but it's partly accessible to the business.

Muddu Sudhakar

>> Very good question, George. So the way in which I see the world is falling on a agents. There'll be some customers who want a single agent to meet all their needs. Is that possible? Absolutely. In some cases, but in the IT world, what people want is, I call them SMEs. They want a subject matter agents for every different use case and what's an agent will be different for you versus me versus John. Once we get that agent, they all need to be in an agent marketplace so that when an employee has a question about software licensing, he can ask only that agent. If you have an issue about a networking issue, you can ask the networking agent. That is my vision is to provide subject matter agents. Now, a super agent can call this agent too, but underneath the agent you need to have workflows because there could be a workflow process, you could have actions, you could have tools, postman collection. Knowledge is just one piece and that is the beauty of what we are doing is we are creating an agent ecosystem. Some agents will be built by Salesforce, some will be built by our partners, some customers will build that ecosystem is what we want to thrive.

George Gilbert

>> But let me unravel a little bit more because it's something that you said that was intriguing, like the root cause analysis one. Like in AI apps, it's a very difficult problem to sort of piece together all the logs, traces, metrics and events to try and determine what caused something. How can you help in that scenario?

Muddu Sudhakar

>> Absolutely. So I'll tell you what we are doing out of the gate at GA. See, when an incident comes to us today in the new product called AgentForce IT service desk. For each incident, we are doing a proactive root cause analysis. How are we doing it? We look at your historical tickets, similar tickets. We are also doing cohort analysis. We are also looking at your knowledge articles, your resolution notes. So out of the gate, the AI can think non-linearly. Humans think linearly. You and me think linearly. I don't know about you and John. John thinks non-linear. I'm a linear guy.

>> No, I'm very linear.

Muddu Sudhakar

>> You're linear, too. There you go.

>> Oh no, I'm very hybrid. But so you say subject matter experts, what we've been talking about earlier all day yesterday was the domain expertise of people and context is greater than quality. If you look at AI, a mixture of experts is one of the core tenets of reasoning.

Muddu Sudhakar

>> That's right.

>> So how do you see this subject matter agents? Is that kind of like an active deterministic subject matter experts? Is this where the action happens? What's your thoughts on that because-

Muddu Sudhakar

>> Very good question, John. I think now you're getting the details of an agent. See, agent by definition should be probabilistic, non-deterministic. If you make that deterministic... in some cases, I have to do deterministic too, don't get me wrong. But agent by definition should be reasoning and inference should be non-deterministic. Underneath, let's say if you're doing a HR agent where it has to do certain processes, you may follow a deterministic process. So some agents may call it deterministic workflow. In that case, you can still call it a deterministic agent, a deterministic-

>> It's situational.

Muddu Sudhakar

>> It's situational. But most customers want probabilistic. The reason people love what has happened since number 2022, ChatGPT and OpenAI has shown the new world of probabilistic. So being non-deterministic, probabilistic is the way to go from a transformation. But at the same time, you have to have the guardrails. You have to have the security and human in the loop. So what we did is, each agent can have the human in the loop guardrails, support the deterministic process, and that's where Salesforce flows comes in. Our MuleSoft flows comes in. So we are leveraging all the properties of Salesforce to build this agent.

>> I know you're new, so you might have an answer for this. You probably do have an answer, but it might not be the official answer from Salesforce. Competition has an advantage, too. How do you see you guys differentiating with the platform in this area? Because again, everyone else has access to the same AI, too. They're probably working on AI. What's your answer to that question?

Muddu Sudhakar

>> I think we have more advantage than the competition is. I'll tell you why it is. First is we have created this system ground up in the last 18 months. Competition has to deal with legacy system. Their data is in disparate system. Nobody has a data cloud like what we are launching, not the competition, not my IT service space. Third, nobody has an Agentforce platform like what we built here. So data cloud, Agentforce. Third, service cloud. The number one jewels of Salesforce is our Serviceforce platform. It's 9-plus billion revenue with 60,000 customers. All I have to do, John, is stay with Salesforce customers, get them on Salesforce Serviceforce platform to IT service. I can be happy every day and get a good growth going.

>> George, we predicted this. This is so much fun because three years ago, we're like, Hey, if they can abstract the way the complexity of Salesforce, and this is like, that's a real advantage. In fact, it's a new category. Ariel Kelman was on earlier and yesterday on the analyst session we were talking about, it's kind of a new animal here. Salesforce has grown into something new.

Muddu Sudhakar

>> Do you like Ariel's and what they did the show? We should give a round of applause to Ariel and the team.

>> Yeah, they did a great show. It's like awesome.

George Gilbert

>> Of course, theCUBE's here.

Muddu Sudhakar

>> And one thing that's a stack, John, to your question earlier, three things we can offer nobody can do, it's the scale, security and the intelligence. That is our difference. That is Salesforce difference that no other competition can deliver today. They're all still dealing with disparate systems designed when I was 25 years old. I'm still 25, by the way. All right? You and me are 25...

>> Are you the agent or are you the real?

Muddu Sudhakar

>> I don't know. I'm younger. I feel more energy now than when I was 25 years back.

>> This is a Cultural Revolution. I was saying this yesterday that this is not the old new guard replacing the old guard. There's a lot of cultural intersections between old and new, senior and young because this is a revolution.

Muddu Sudhakar

>> It is. It is.

>> And that's not talked about a lot, that it's not just like the young guns. We're coming out of the woodwork. We love this market. Come on, it's the most exciting thing to us.

Muddu Sudhakar

>> I call it the cross-pollination. It is people who know about AI and don't about AI as said, young, old, the mix of everything, and this is what I like, the beauty about this, and no human being should be left untouched.

>> It's like us, my old guys like, Hey, we used to walk to school in the snow with bare feet. Now it's a wonderful time, right? Because we knew what AI could do, but it never was ready. The infrastructure, the timing of the infrastructure, AI infrastructure, growth, the platform layers, the harmonization layers that George documents, it's all kind of in place. The pieces are there. The question that we have and George has is how does this evolve and scale up to create faster value? The data challenges changes all architecture. So there's a lot of work. It's a very nuanced point, but if you look at the companies that George interviews, they're challenged to re-architect for the next generation. Let's look at 20 year cycle.

Muddu Sudhakar

>> And by the way, that's where it's unsung hero for us is our data cloud. It's beautiful. Like the knowledge articles could be in multiple systems. It could be in SharePoint, it could be in Confluence, it could be in some other system, bringing all the data and the resolution notes. That's the beauty.

George Gilbert

>> Okay. This is my question. So tell us on the HR product, where you're getting from your systems of record and then how you're adding new context to that. But then for the IT service product, what systems are feeding it and what context you're adding?

Muddu Sudhakar

>> Very good question. Let me start with HR. By the, I run HR service also. So what we are doing for HR is, we are giving you that system of engagement, system of experience, and system of work. So being Slack, our Teams, we are giving you that experience. Under that, there could be a system of record from Workday, it could be SAP SuccessFactor, Oracle HCM. It could be ADP, Paychex, Intuit, Kronos. There could be hundreds of systems. Typical HR company will have usually 5, 10 systems. But what they want is one system of experience where the HR question is about my payroll, about my benefits. It's about changing my address or adding a dependent. They want one place to do it in a nice experience manner, not create a ticket, wait endlessly. So that is what we want to solve in HR experience.

>> And then on the IT service side.

Muddu Sudhakar

>> On the IT service side, what we have done is not only we want to care about the experience... again, the most important thing, we want to meet users where they are. Slack first, Teams ready, both are there. But we are also launching the service desk. That's our system of record. That is for service management, incident, problem, change, release, service catalog, then service workflows. All of that is in the service desk for us. That's our system of record. Then in the service record also, we are trying to do agent assist classification, routing, assignment, summarization, like John talked about root cause analysis, provide to incident resolution. All of that is done in the system of record. Then fourth area, I'm launching CMDB agentically, which is through discovery, service map, service graph.

George Gilbert

>> So the agent would discover the graph of the dependencies.

Muddu Sudhakar

>> Like we used to talk about in the past, the good old CMDB result. Now I'm doing the CMDB in the new age. We are the only company which can do software-based discovery and agentless discovery in an agentic manner. Put them into a latest CMDB on Salesforce platform and provide your service graph of all your infrastructure. It's similar to what we talked in the past of your apps, services and your assets.

George Gilbert

>> And that goes from the app layer all the way down to the infrastructure?

Muddu Sudhakar

>> Infrastructure. And Kubernetes layer too, cloud infrastructure.

George Gilbert

>> Okay. Because if you can build a rich enough model of that, then you really can do diagnostic analytics, and then if you have high confidence in the diagnostics, you can recommend or even eventually automate some remediation.

Muddu Sudhakar

>> That's exactly what we're doing, John. If not, that gives you the context of the organization-

George Gilbert

>> But then-

Muddu Sudhakar

>> The root cause analysis-

George Gilbert

>> What is your secret sauce in discovering that graph? That's the foundation of all that?

Muddu Sudhakar

>> How do you discover the graph? How do you make your CMDB dynamic? How do you make it ephemeral? How do you know your assets are correct? How do you know the metadata with the CIs? All of that.

George Gilbert

>> It's that secret sauce because that's a high value asset.

Muddu Sudhakar

>> It's a lot of secret, secret sauce in models, it's correlation, root cause analysis, doing the models to interpret those things. All the intelligence around the graph, connectivity, physical connectivity, virtually all of that we built in.

George Gilbert

>> So then you would be in a position to really be like the operational backbone for not just the customer 360 or what's now, I guess Agentforce 360 apps, but the custom apps that extend that.

Muddu Sudhakar

>> That's the goal. So right now we are launching, as I said, CMDB discovery, everything. As we go into outside of the two operational intelligence, AI ops and AI observability, we will help the customer 360.

George Gilbert

>> Oh, now I'm getting the picture because the data cloud guys want to go beyond customer 360. They want to extend into the operational domain, and then you'll have the manageability behind it.

Muddu Sudhakar

>> That's right. You got it.

George Gilbert

>> Okay, okay-

>> Very good. Pretty powerful. Yeah. So first of all, again, we'd love... again, we think it's a whole nother category. So I want to get, again, you've had many chats about this, me and you on theCUBE. What's next? You have a ticket to market. I still think Salesforce should be called agentforce as a company. I think the name Salesforce is too limiting. It's gone beyond that. But now you guys get into operational, so we're on a road to super intelligence. Okay, put that aside. We're not going to debate the year, but it is highly accelerated right now. Things are moving super-fast. What are you going to do as the leader to get the word out? What's your go-to market strategy? How do you tell the story? Because it's not the old story, it's a bigger story.

Muddu Sudhakar

>> By the way, when you said super intelligence, that reminds me of the super cloud that we talked about.

>> Exactly.

Muddu Sudhakar

>> We should do that. Super agent, super cloud.

>> Mega cloud, super cloud, gigawatts, racks. We are in a massive Cambridge explosion. But the bigger winning too, it's an incumbent company. Salesforce is at scale.

Muddu Sudhakar

>> Okay. So I think, look, my focus right now as the go-to market side is win customers, as many customers, and again, focus is start with Salesforce customers, whether they're SMBs, mid-market, commercial, enterprise. We want to focus on Salesforce customers as much as possible. Go after them and particularly service cloud customers and extend their customer service into IT service. We want to win them, deploy them, get customer references. We are launching with four customers as a reference at the launch. We have Feedmont, we have Aspel, EPB, right, we have UNESCO. They're going to be at my keynote. Please tell your viewers to come to my keynote this afternoon at 3:30.

>> Okay, 3:30, hear that.

Muddu Sudhakar

>> Intercontinental, please come there. And then look-

>> Where Intercontinental Hotel?

Muddu Sudhakar

>> Intercontinental 3:30 today. That's a keynote for IT service. But look, there's more to come. We have to do enterprise service management, IT asset management, GRC, do more on expanding the CMDB discovery. Grab into all the infrastructure side.

>> What do you say to a customer that has the competition or someone who hasn't innovated, they have IT service management, they're managing all their tickets and all their platform legacy systems? What's your message to them? How do they onboard the Salesforce? What's the path? What's the message to them? Why and how do they do it? What's the three steps they need to do?

Muddu Sudhakar

>> I have three things to offer. First, take a one-hour meeting with me. Understand what we built, product deep dive with my team. We spent thousands of resources to create this, at least give us one hour chance. If you like it, we do something called light up the org, John. Light up the org is we take their data, sample data, anonymize, load up into our org with Salesforce instances fully created... So I'm going to create a... before you buy the product, you'll see your incidents, your root cause, your management, everything with us. That's right. Third, if you like it, then we want to co-partner, co-create with you, your use cases, and work with you as a partner.

>> Yeah, and you're going to get the word out. Thank you for coming on theCUBE. Congratulations on the new role as the senior leader at Salesforce, senior vice president and GM. A little P&L pressure. You're going to get those numbers up, IT and HR service at Salesforce. Again, the world's changing, and you're in the driver's seat. It's a fun time. Thanks for coming on.

Muddu Sudhakar

>> Thank you, John. You've been a good friend for me. Thank you so much. Thank you, all right.

>> George, thank you.

George Gilbert

>> Thank you, Moodhu.

>> For George Gilbert, and I'm John Furrier, host of theCUBE, we're live-streaming here in San Francisco, California at Dreamforce 2025. We'll be right back after this break.