The Autonomous Service Workforce: How AI Is Reshaping Customer Operations

Incremental automation has reached the limits of what it can deliver for customer service organizations. Enterprises that spent the last decade improving ticket routing, reducing handle times, and layering chatbots onto existing workflows are now asking a fundamentally different question: what does service look like when AI can operate autonomously across the full resolution journey?

Keith Kirkpatrick and Melody Brue sat down with Tom Eggemeier, CEO of Zendesk at Zendesk Relate 2026 in Denver, to unpack what the shift toward an autonomous service workforce actually means for enterprise CX strategy. The conversation examines why specialized AI agents are replacing generalist automation models, how resolution-based pricing is changing the economics of service, and what organizations must do now to prepare for an increasingly agentic operating environment.

Tom also shares where enterprises are running into challenges as AI expands beyond isolated pilots and into larger customer and employee service environments, along with his perspective on what separates the companies successfully scaling AI from the ones struggling to keep pace.

‍

Key Takeaways:

🔹 Service organizations have moved beyond task automation. The market has shifted from asking how to automate tasks to asking how to build service operations that can resolve issues autonomously from end to end. Zendesk's autonomous service workforce thesis is a response to that structural change.

🔹 Specialized AI agents outperform broad automation. Context, policy, and workflow-specific reasoning matter. Service operations that require nuanced judgment for specific domains, not broad-purpose tools deployed across every interaction type.

🔹 Resolution-based pricing is reshaping the service platform market. As AI operationalizes across service environments, outcome-based commercial models are replacing seat-based pricing, shifting value measurement from activity to results, and redefining how platforms justify their cost to buyers.

🔹 Scaling AI exposes operational complexity fast. Enterprises expanding AI across customer and employee operations are encountering integration overhead, governance gaps, and organizational change management challenges that were not visible at the pilot stage.

🔹 Human expertise doesn’t disappear in autonomous service. As AI handles higher volumes of routine resolution, human agents shift toward higher-complexity interactions where judgment, empathy, and contextual reasoning cannot be replicated at scale.

🔹The enterprises that lead the next phase of CX transformation will not be the ones that automated the most. They will be the ones that built the operational architecture to sustain autonomous service at scale without sacrificing the human judgment that still matters most.

Watch the full video at sixfivemedia.com, and be sure to subscribe to our YouTube channel so you never miss an episode.

‍Disclaimer: Six Five Media is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded, and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors, and we ask that you do not treat us as such.

TOM EGGEMEIER:

People sometimes say I got to do a two to three year project or a 12 to 18 month project on cleansing my data, then I can start with AI. We say take your data as is, do a little cleansing and start with those one or two or three use cases and you're going to see the power of that.

‍

KEITH KIRKPATRICK: 

Hello, and welcome to Six Five On The Road at Zendesk Relate 2026 here in Denver. I'm Keith Kirkpatrick, joined by Melody Brew, and Tom Eggemeier, CEO of Zendesk. Today, we're going to be talking about the company's vision for an autonomous service workforce, the evolution of AI-powered service platforms, and what this next phase of operational transformation means for enterprises.

‍

MELODY BRUE: 

Big day for you, Tom, huh? You got off the keynote stage. You've been going meeting to meeting. What's your day been like?

‍

TOM EGGEMEIER:

 It's great. I get to see customers, and usually I have to travel around to see them, so 30-minute increments. I've seen, you know, half a dozen, probably a dozen customers by the end of the day, but I'm always excited about these days.

‍

KEITH KIRKPATRICK: 

Yeah. So let's talk a little bit about the keynote. In it, you really frame the industry as moving beyond software to this autonomous service workforce. What's happening in the market that really is driving this change?

‍

TOM EGGEMEIER: 

I think a couple of things. One, I know AI is the buzzword everywhere, but I really do think we've moved from conversational AI to a genetic AI. and starting to conceive of AI agents as work and doing tasks. And people are really starting to think about how they evolve human beings and AI agents in solving problems, including complex problems. And I think we're just at a stage now where AI is being accepted. People are thinking about how to systematize this and to really get great results out of this new workforce.

‍

MELODY BRUE: 

Yes, systematizing that, it really is an operational difference in how the enterprise needs to look at the service. How do you think that customers, your customers, should be organizing around that new operating model and where that provides value to their customers?

‍

TOM EGGEMEIER: 

So first, I think what I've seen, and we've gone through these things like trial and error, you need to pick three to five use cases that you think are going to really impact your customers and your employees in a positive way. We started with a thousand flowers are going to bloom within Zendesk. There's still some blooming flowers, but we found better to a little bit go bottoms up and top down in a couple of key initiatives. That's number one. Number two is really have clear goals on what you're trying to accomplish with automation. I tell everyone automation and efficiency is not enough. You need to do automation and drive for customer satisfaction or employee satisfaction because automation is not an end in itself. You're trying to serve your customers at the end of the day. The third thing is technology is not enough, even though we sell AI and software solutions. If you're not going to go change your processes and your workflows and reskill your teams, I would tell people don't invest in the technology.

‍

KEITH KIRKPATRICK: 

You know, it's interesting you mentioned that it is about people and processes and technology. But one of the things that I think is often overlooked is sort of the foundational element, which is data. So I'm curious about how is Zendesk really helping to deliver additional context and insights that can actually drive better customer experiences and ultimately business outcomes?

‍

TOM EGGEMEIER: 

So Keith, we're doing a couple of things. Number one, we have over 18 billion customer interactions and employee interactions that are kind of old school rated thumbs up or thumbs down. And that data set is really good for, you know, thinking about how we improve customer outcomes. The second thing is that we really recommend, everyone's going to cleanse their data, but people sometimes say, I got to do a two to three year project or a 12 to 18 month project on cleansing my data, then I can start with AI. We say, take your data as is, do a little cleansing and start with those one or two or three use cases, and you're going to see the power of that. Third, we're really big on Knowledge Graph, and we launched a bunch of things today where we are interconnecting with all kinds of systems, all kinds of data, because we think the richer and more context you have with your data, the better outcomes you're gonna get. So those are the three things that we talk to customers about.

‍

MELODY BRUE: 

Let's talk about the economics a bit. AI has changed the way things are priced, but it's also changed the way that we look at value. As you shift to this new operational model and a new pricing model, what are you looking at and how do you think that this will land or how is it so far landing with your customers?

‍

TOM EGGEMEIER: 

So far, it's been really positive. We launched outcome-based pricing for automated resolutions, I think it was two to three years ago now. And at the time, we had a little trepidation of changing our business model. It was one of the few times I say, it's great being private because we can experiment things like that. You know, if something goes wrong, it's not going to be big public news. But customers have overall really, really liked it because we're in the same boat with them. We are trying to drive an outcome. I tell everyone, first thing I do in the morning is I look at how we're doing with our customers on automated resolution. So I get a daily snapshot of that. And with some of our latest tools, I can drill down. Why is it up? Why is it down? And what we're doing about it. And so I think the whole software world is going to go from seats to consumption to outcome-based pricing. I think it's a lot easier in service, because everyone kind of knows, was it solved correctly or not?

‍

KEITH KIRKPATRICK: 

What's the outcome? I'm just curious though, you work with a lot of different types of customers, large and increasingly larger customers. What do you think are the sort of traits or behaviors or actions that separate those that are actually successfully operationalizing AI compared with the ones you don't?

‍

TOM EGGEMEIER: 

I think it's a real ownership, and I don't just say this as a CEO, but I think the CEO matters here a lot, where the CEO really needs to get embedded and really understand AI, really understand what you're driving, and really drive it within the organization. I think this is one of those where you have to lead from the front. I am a history major with a law degree, not a software engineer, but I've spent a ton of time the last years, and particularly the last six months, using whether it's Codex or Cursor or Clod, really truly trying to use the tools myself, building AI agents, actually shipping some code a couple weeks ago, doing a project with my son on creating an application, because I think you really need to understand, because I understand the good and the bad, and some of the pitfalls better than I would without using the tools.

‍

MELODY BRUE: 

I think it's important also to take the view of the consumer, not just the customer, but the customer's customer, right? So you guys actually have a bit of a layer of who you have to think of, where's the resolution and where does it get answered? And so does that help you as a consumer to kind of view that from that perspective of like you're not just, we're not just talking about software. We're talking about people's travel plans. We're talking about their job, what their benefits are, whatever it is. It's more than just software that we're talking about.

‍

TOM EGGEMEIER: 

Totally. We have about 60% of our business being business-to-consumer, 40% business-to-business. And then we have another part that's business-to-employee. So it obviously doesn't add up to 100%. The 60% and 40% are 80% of our business. So we'd have to go multiply those by 80%. 20% is business-to-employee. I take it really seriously that we have the ability, particularly now with AI, to delight our customers' customers. Consumers are always the thing that we can all relate to, Melody can relate to. A lot of times when it comes to those emotional moments where easy transactions are easy, you can go automate those, but a lot of times it is a real key moment in someone's life and you can influence and hopefully give them a point of magic, a point of delight. We do talk about not just the companies we serve, but kind of the three audiences, consumers, businesses, and employees, and how we're going to have to do things slightly different for all three audiences.

‍

MELODY BRUE: 

Yeah. Well, that's a great way to end this week. Thank you so much for joining us. It's been a great first day here, and we have lots more to come. And thank you for tuning in to The Six Five On The Road at Zendesk Relate 2026. Don't forget to hit subscribe, like, and share on socials, and check out all of our coverage at sixfivemedia.com. Thanks for joining us. We'll see you next time.

‍