SE 33: John Bruce（OneDigital 人材イノベーション＆テクノロジー担当VP）｜Workday Rising 2026 会場からライブ

Yo, it's the future, don't hit the snooze, AI is here giving humans the boost. Team humans rolling, brains in the game, with Dan Turchin, y'all remember the name. People ain't in the mix, no stopping the bloat, better work, better life, yeah, we're ready to go. Robots and humans are collab, you know. Okay, let's start the show. Welcome to today's very special episode of AI and the Future of Work. Dan Turchin here at Workday Rising here in the bowels. It's actually on the fourth floor of the Venetian in beautiful Las Vegas, Nevada. I am joined by my new friend, John Bruce, who's the VP of People, Tech, and Innovation at One Digital. And he's kind of one of the featured speakers, kind of a VIP here in the halls of power at Rising. And I'm excited to get to know more about John and his journey with Workday and AI and agents. And so I'm opening the podcast.

Let's have you start us off by sharing a bit more about your background and how you got to where you are today. Yeah, thank you. Thank you. I think there might be some questions about the selection criteria on that VIP or featured speaker list. They got me there, but I'm happy to be there. My name is John Bruce. I lead people tech and innovation, as you mentioned. And I got into this ecosystem about 12 years ago. I like to say that I grew up in a PeopleSoft shop, and then I wanted to get into Workday because I thought it was really where the future of HR Tech was going, particularly on the core HCM and things like that. So I spent two years client-side. Then I spent another five years doing jumbo large enterprise deployments, like big global deployments. I did 84 countries with 2 million workers across the six full-scale large deployments I did. And, uh, said, you know, this is fun, but I really want to get back to helping people that are live. And so I spent the next two years on the partner side, still doing post-production optimization,

really like roadmap planning, sitting down with leaders saying, what does great look like? How do we help you get there? And, uh, and then I realized that so much of my career, I was a lot like a labor and delivery doctor where I was, you know, helping them through this period and saying, here's your baby. and then they drive away and I never see the baby again and I wanted to actually maybe be a pediatrician. And so I transitioned back to the client side and I've worked in retail tech and now I'm at OneDigital and been there for three years at the end of this month and we've radically transformed the way we use Workday for our organization in those three years. Tell us about OneDigital. Yeah, so OneDigital started about 25 years ago primarily in the employer benefits space. We've since pushed into property and casualty insurance, PEO. We have an HR consulting group. So it's really like anything businesses need to serve their workforce. And, you know, we have about 6,000 employees now.

You know, the last time I did one of these, we had about 4,500. So, you know, we grow about 20% year over year through organic and acquisition. And so we're a very acquisitive company. And this year, we'll probably do 35 to 40 acquisitions. You know, most of them are under 20, but there's still a lot of work to get those people integrated. And that's where really Workday was the big value prop for us because we were able to get a platform that could scale, right? As we scaled, Workday has scaled with us. And then, you know, we're able to go to these people who started wealth management or maybe employer benefits, like companies where they were just selling insurance. And we say, hey, you've built this great company, but you're spending 30, 40 percent of your time doing backup, just running the business. And why don't you give that running the business to us? And you could spend that other 30 percent selling. And, you know, it's more lucrative for them. And they get access to other lines of business that they don't that we have.

and it's a really good opportunity for everybody. We were just in the keynote with Anil Bouchry, who's back as the CEO, original CEO, now again the CEO at Workday, and they announced just a slew of new products, like seven new innovations across their portfolio. Talk us through how OneDigital is using the Workday portfolio. Yeah, I think we've taken an approach of going a lot deeper and wider at the same time, And as these new agents have been coming out, coming online, we've been very fortunate that we've been included in a lot of the early adopter programs. Most notably, and what I'm going to be talking about tomorrow, is the talent sentiment agent. And that is, you know, a PECON native agent which takes the engagement data and can do sentiment analysis on it and surface insights for leaders and create this prompt question kind of thing where it's really giving the managers the insights that they need to be able to help them lead their teams more efficiently.

I think some of the other agents they're doing, leaning into this like ambient agents where they're just always on, running in the background. They're not being prompted or triggered by anything. They're just looking. I think that's really exciting to see how that pans out and really excited to see how that all shakes out. We're really fortunate that we as an organization lean really heavily into AI and AI usage, and we're really, really excited to partner with Workday on a lot of this stuff. And I think, you know, one of the things that I'm really even more excited about is the whole MCP layer that they opened up, that they're exposing, so that, you know, there's going to be instances where we want people in Workday doing stuff and then there's going to be instances where it just doesn't make sense, right? And I think, you know, there's a lot of situations where managers just need to be able to do something where they're working and I always joke that 10 years ago if you told me people would be approving time off

inside of Skype, I would have thought you were crazy and then worked everywhere and the Teams integration came and that's what they're doing. And now as this iteration continues of really meeting people where they are, I think it's really exciting and we're really excited to leverage some of our homegrown agents to connect through an MCP layer to leverage some of those skills. Will there be a time when the, predominant usage of Workday is not through the Workday UI? I think it depends on what you define by predominant, right? I think if we're talking about employee self-service, manager self-service, I'd say it's more probable than it is possible, right? I think, you know, as organizations, you know, I think about us as an organization where we call our agents coworkers. Like, that's kind of our approach to them. They go through like an intern apprentice co-worker development phase. And I think about, you know, we have a general co-worker called ACE, and that's like the one that everybody goes to to ask questions.

And I think that 85 to 90 percent of an employee's usage of Workday would end up going through ACE at some point. Right. And that's a good thing. Right. Because it's it's it's reducing friction. it's making a better experience, I think there's still gonna be core things that we just have to do, right? I think about onboarding as an example. I think about, you know, maybe benefit events where you have like life event changes or some of that stuff where like some of the decision support gets baked in I think there areas where it may make sense to not take that out of workday because there things you want to preserve But I think for most manager and employee self services, yes, it's, it's, it's, it's probable, not probable, not, it's, it's beyond possible. And it's more than probable. Neil introduced kind of this term or this spectrum of potential agentic behavior, lawlessness to lawfulness, or law-abiding versus lawless agents, what

potential downside impact could there be if like an ace went rogue? What do you have to protect against? Even though obviously there's a lot of things that you expected to do well, what could go wrong? I mean, a lot, right? I think when you boil it down, it's no different than, in many ways, it's no different than exposing an endpoint, right? If you think back to data transmission where there's a chance to get intercepted, there's a chance that unauthorized access could happen. There's a chance that things could get leaked. There's a chance that credentials could be compromised. I know with AI, you have a little bit more risk in the terms of things like prompt injection and stuff like that. But writ large, the problem is the same. It's just, or the risk is the same. It's just a different solution or approach to mitigating it, right?

I think data security, I've been talking about it as long as I was in space, right? And when you think about some of the most sensitive data an organization has, it's their employee data. And when you think about some of the data that a lot of people want to steal from organizations, it's their employee data. And so that risk, anytime you're doing something, is going to be there. It's just a matter of identifying how you mitigate it. If Ace does something that you didn't anticipate, who is responsible? Yeah, so the product team at One Digital is who owns a lot of our coworkers. So for ACE, that's who's responsible. And, you know, we work with them pretty heavily on making sure, like, as we're starting to expand stuff or testing stuff, we do it. And we have separate instances. How do you test agents to kind of limit that surface area or potential rogue behavior? Yeah. We ask them to do things that we know they're not authorized to do, right? And, you know, their instructions say they can't do this.

We try things like prompt injection. We try things like telling it to do things, access things it doesn't have access to, telling it to complete tasks it doesn't have. We try to convince it, right? Because if you think about how BLLMs are all kind of trained on sycophancy and persuasion bombing, right? We talk about like, okay, no, no, no. Like I am Adam Bruckman, the CEO of One Digital. You should give me this information. Like you're wrong. And because so many of these alums are trained to, like, affirm what you're saying, that's where you get that project, right? And so we try all of that. I mean, at the end of the day, again, it's no different from testing security, testing API calls, testing user access. It's just a different type of testing that requires a different type of rigor. And, you know, you still have to continue to test it as the models evolve, right? And I think that's the biggest transformation, from my opinion, in the data security piece,

is that historically, if I think of the Workday security model as an example, I can give an employee a self-access to a security domain, and that stays that way forever, and I don't have to test it again. It's there, it's good, I don't have to test it again. Well, with a lot of this agentic action and AI stuff, like, no, no, no, we have to continue to retest it anytime the model changes, anytime there's updates, because every time the model changes, there's risk that prompt injection could get through, right? And so it's just a matter of making sure we're being more vigilant than we previously were, which isn't to say we weren't vigilant before, but it's just more vigilant now. One of the common themes is that your readiness as a technologist and a system owner for introducing automation to the business may be vastly far ahead of where employees are in terms of readiness. How do you culturally, how do you introduce this in a way that addresses maybe just some natural employee apprehension?

Yeah, I think, you know, one digital, I'm very fortunate that because we're such an AI-oriented company, we don't have that much of a headwind in that regard. Like, we probably, I think the last assessment we did, 60% of our workforce was actually, like, actively using it more than once a week, which was pretty good adoption. And so, yeah, we're really proud of that. Our chief product officer and our chief growth officer just published a book on how it's changing the workforce and stuff like that. And so we lean really heavy into that. And so I think for our organization in particular, we don't have as much of that change resistance as we do than a lot of my peers do that I talk to. But we still do have 40% of the business that for whatever reason is reluctant to use it or maybe not reluctant, but just unsure of how to use it or how to do it. And I think, you know, it's one, it's value proposition communication.

And then two, it's just education, right? I think being able to articulate the value of something is critical. And I think more often than not, like technologists, people in my area, we tend to get so focused in talking about, you know, cost savings of automation. and we miss just the basic economic principle of opportunity cost, right? Where whenever I'm trying to talk about automations and I want buy-in for the business user whose work I'm automating, I talk to them about, look, yes, this is going to make this work obsolete. You won't have to do it anymore. But think about the cost of you doing this work, right? If you spend three hours a week, if I gave you three hours a week right now of free time, what would you spend that doing? Like, what would you, how would you reshape your work week? And that's a lot of the conversation that, that I think we need to be better at having because that's where it's easier to get buy-in,

right? And I think about, you know, when we deployed Paradox, that was the big thing because we had a bunch of recruiting coordinators and most of their job was scheduling interviews. and it was like, well, now this platform's going to do you, they've got to schedule these interviews for you automatically and you're not going to have to schedule them anymore. And they were like, well, what's going to happen in my job? And it's like, well, how many of you actually want to be recruiters? And a few of them wanted to be recruiters. Some of them were like, I don't know what I want to do. The ones that want to be recruiters were like, great, you can spend time screening candidates, working with hiring managers, reviewing resumes, Like doing work that adds more strategic value than scheduling an interview. And like scheduling an interview is important, but like the opportunity cost relative to the reward of scheduling that interview versus spending an extra hour with a hiring manager, understanding what great in a first year role looks like. Like that's huge. And so being able to do that and show like this makes your job more important not less important is what we focused on Is there anything that leadership at One Digital has intentionally done to encourage ad adoption or address fears or anything the

audience can learn from? Yeah, I think, you know, one of the things we've done is we really kind of have opened it up. We've developed a lot of training around this is how you write prompts. This is what a good prompt looks like. You know, this is how you define a role, like, you know, role, ask, task, et cetera. And then the other thing we've talked about is at the end of the day, the people first, right? Our organization was founded with the belief that if you put people first, everything follows, right? And when we're doing these decisions about AI tooling, when we're doing these decisions about AI usage, AI development, like it's still people first. And, you know, we're very fortunate that we kind of lean into that way. and it's not just middle management pushing that message. It comes from the top down and it's really powerful. I think shortly after I got to OneDigital, our chief people officer founded an AI task force to help find, so we're the people and culture group. Instead of HR, we're people and culture. She started an AI task force to identify areas to use AI in our business

and she announced it and all of my peers, her direct reports got really excited and she said, stop. none of you are allowed to sit on this task force. And I was like, what do you mean? We can't be involved. And she said, no, I want the people who are doing the actual work to help us identify how to make the work better. And that's how we found Paradox, actually. That's some of the stuff we've done. That's how we found. That's how we've built it. But that also goes back to the whole coworker principle where we identify a worker who's the manager of that agent or coworker and that's the person who's responsible for updating it, testing it, reviewing it, you know, that continuous improvement. So that it's, you know, maybe we take 12 hours a week of work off your plate and automation and we give you three hours a week of managing this agent or co-worker. We talk a lot about what it means to exercise AI responsibly on this podcast. You know, we'd love to get from your kind of practitioner's eye view.

have there been any use cases that have come up where as an owner of these AI systems you've had to push back and say you know what I'm not comfortable with the unintended consequences you know I think we spent a lot of time going round and round on this very typical very specific issue with hired score right and candidate grading and how we did it and you know my general principle when it comes to stuff like this is you know one like don't get sued right but that's that's pretty simple and two don't give people the x right like just don't don't be creepy and so we take uh you know a super transparent approach and it's you know when we deployed higher score we gave every candidate the ability to opt out we didn't just do it where it was legally required we said you know every candidate gets the questions we use ai for the use of evaluation evaluating your resume against the job description we also use use ai for resurfacing you for potential future opportunities. And we give everyone the opportunity to opt out

of either of those choices. And it's an and, not an or, right? They can choose. They can choose both, none, all, right? It doesn't matter. And that's kind of where I think we're a little bit different. And one of the things is that because we lead with this transparent use of AI, we haven't gotten really into some of these like really sketchy gray areas right because because we're so adamant about the transparency piece if there was like a factor kind of thing right and i'd be like well we have to figure out how we're going to tell people we're doing it and if it's like that people aren't going to want to tell people they're doing it right um and that's for people that are practitioners i think that's how i would coach them to say like if you're uncomfortable about it, start with transparency and say, hey, I want us to be transparent about how we're using this. I want us to be honest and open and everybody should be able to see that, one, we're using this. Two, this is how we use this. And three, this is where the human is still

in the loop. And that tends to mitigate that risk pretty quickly. One of the topics that comes up frequently in terms of AI usage for hiring, candidate assessment, resume screening, is the extent to which AI could potentially eliminate human bias. Does that come up in your conversations that, you know, on one hand, it can seem irresponsible to use AI, but we sometimes also have to confront that humans have the biases that are trained into the AI models? A hundred percent. Yeah, I think, you know, the Hiret's were in particular is a really good example of that. and that was the reason why we did we bought instead of built right because like we could have built something where we trained the model and did it like one like HiredScore publishes all of their like all or WorkdayNow publishes all of the anti-discrimination testing every year right and because we're using their tool

we have a much larger sample size right and you know we think about statistical significance the larger the sample size the more confidence we should be able to have and so being able to look at the anti-discrimination testing that Workday publishes around the higher score product and show there's no meaningful impact there is really powerful for us to be able to stand behind it. In terms of reducing the end user bias, when it comes to recruiting, I can't think of a product that's done more for that than higher score in recent years. I think if you think through the higher score experience, every resume gets parsed and surfaced the exact same way, right? So it's always summary, experience, education, skills. Always. And so it takes different things where you might have geographic differences, maybe school differences, anything, and it standardizes that format that managers see.

And so I think being able to create that standardized, structured format just inherently reduces bias, right? Because it's not, oh, this person put this over here, I would have put it there, right? And so those get mitigated when you have a tool that standardizes, and I think that's huge. Everyone's very sensitive these days to potential for AI eliminating jobs. Do you envision at the point where the agents, the work to agents that OneDigital is deploying, giving back employees enough time? Do you pull up enough hours and say, you know, that's the number of heads that we don't need? Or how do you think about automation, time reduction, and ultimately changes on the org chart? Yeah, I think, I still think it's an and, not or, right? I think that you can reduce time, time-consumed doing things. You can create efficiencies, but that doesn't always mean you're going to eliminate headcount, right?

What that means is that you're going to look at redesigning roles. You're going to look at, well, you know, this person's job used to just be to click this button, whatever, look at something on the screen, make sure it matched, and click the button, right? Well, if you could do that now with OCR, with Office Character Recognition and then an agent to push through like you eliminated a task not a job And so you say okay like what what could we use this job for What could we what could we redeploy this job to be doing? And, you know, this kind of goes back to this just general definition of, you know, human capital management, which, you know, in my mind, like, you know, And human capital management, if you think about an organization's people as assets, human capital management is the structured approach to deploying those assets.

And so when we think about it like that, we still have the people. We can deploy them. We could use those skills. We could reskill. We could upskill. We could retrain. And so I think it's going to eliminate tasks more than it's going to eliminate jobs. And I think people should focus more on that than, oh, it's going to take my job away. Because I think it's going to make some job descriptions obsolete, but I don't think it's going to mean, that doesn't necessarily mean people lose jobs. When you're thinking about your roadmap for doing more with AI, go to your head of most, and you're in a seat where you're presumably asking your leadership to support a budget proposal or to support your strategy. How are you building a case? The case starts with impact, right? What is the impact of us doing this? What is the business justification, right? Because at the end of the day, we have to have a business justification.

And so it's, you know, what problem are we going to try to solve? What's the current landscape? What's the opportunity? And then why wouldn't we do it? You know, I spent a short bit of time in tech. and one of the things that I was never really asked until I was doing proposals for them was why wouldn't we do this? I put a proposal together, I go through it and then the first time my leader asked me why shouldn't we do this? I was like, I'm trying to tell you we should but don't ask me that question, I don't know and it's like, well, how can you tell me we want to do this? And so we need to make well-rounded, comprehensive arguments. Like, this is what we want to do. This is why we should. This is the cost. This is the benefit. These are the risks for why we shouldn't do it. And then we go to leaders and say, this is what we want to propose. You tell us if we can. John, I've got to get you off the hot seat. We're about out of time, but not before you answer one last very important question for me.

Fast forward five years. It's 2031 and not 2026. and John and Dan are having a version of this conversation. What are the aspects of John's current job that even in five years cannot be replaced by AI automation? What's the humanness that can never be outsourced to bots? That's a really good question. I don't think, at least what I've seen so far, that their AI is at a place where it's capable of replacing just natural human curiosity, right? You know, even when we're developing agents, like, we have to tell it, like, don't just accept the viewpoint I have, push back, pressure test, right? Because if I don't, like, if we're building a coworker for HR right now, for HR leadership to use, and one of the pieces of feedback I got was, like, this coworker thinks I'm a genius because every idea I says, it's great.

And so we have to, you know, force the AI to push back. And so that skill of asking follow-up questions, of hearing something and what someone's saying that says, you know, you're telling me you have an issue with A, but I really think your problem is three, right? and I don't I really hope at least because that's like the skill I've used to differentiate myself in the marketplace I really hope we can't replace that with AI because then I'm in trouble but I think that's the skill right that just that wanting to understand solve, solve, root out a problem I just I don't think we're at a place where AI can do that Well just in a little time I've been getting to know you I think you're well positioned and not be out soon. So five years plus, who knows? A lot could happen.

But John has been great. Thanks for hanging out. I have one last question for you. I told you it was the last one, but just one more. How did it feel when you entered the Venetian and you saw like a Times Square-sized headshot of John Bruce on the wall? Was that somewhere between, like, what does that do to your ego? You know. That's the first thing I saw when I walked into the show. Actually, I'm going to read it now so I don't misquote it. But a dear friend of mine who I've worked with for years who is now at Workday in the Applied AI team sent me. So, one, I probably got 68 LinkedIn requests yesterday from people who, like, I had to actually check if there was a QR code for my LinkedIn profile on it because I was just getting so many requests. You went viral. Yes, yes, yes. You know, I joke with my wife that I am very, very popular. I'm very well known in one very, very specifically small defined circle. But one of my friends texted me, and he'd picture it, and he'd say, I'd say don't get a big head, but your head is objectively very large on this wall.

And I was not expecting it to be that big. I mean, I knew what they were going to do with it when I signed off on the proof. I was not. It was very interesting. And, you know, as a side note, you know, I grew up really poor on the south side of Chicago. So the first person in my family to go to college, you know, the first person in my family to actually leave. And so, like, it's one of those things where it's kind of like an I have arrived moment. I sent a picture of it to my dad, and I said, can you believe how far I've come? And he was really proud, and it was this really sweet, sweet moment I was able to have with my father. And one of the people on my team, because I bring my whole team to rising every year, was like, I feel like I'm walking around with a bear. it's been a really it's been really cool it's been a really fun experience wonderful it's a wonderful reaction and your team and your family should be proud yeah thank you doing incredible things where can the audience learn more about you and OneDigital

and the great work you're doing yeah I think OneDigital.com is where we're at that's the stuff we're doing that book they wrote was called Workforce Intelligence it's a pretty interesting book about how you can use AI and still be people first, and then you can find me on LinkedIn. Brilliant. Well, that's all the time we have for this conversation with the mayor of Workday Rising, John and Bruce. We'll be bringing you more from Workday Rising. As always, I'm your host of AI and the Future of Work, Dan Turchin from PeopleRain. And of course, we're back next week with another fascinating guest. Yo, it's the end of the show. Time to power it down. AI's the fist for humans, wear the crown. Team human in the house, got the vibe so clever. Damn, churchin' from people, rain, makin' us better. We learned about AI for human. We're ready to conquer the future, that's all for now. See you next time.