Executive Summary:
Discover how is revolutionizing HR through AI in this
episode of The Enterprise Edge Podcast, recorded live
from Oracle AI World 2025 in Las Vegas. Join host Mark
Vigoroso and Hubert Winter, Senior Director of HR
Technology at Quest Diagnostics, as they explore the
cutting edge of workplace automation and human
capital management. From piloting Oracle's new Manager Concierge AI agent to dramatically reducing help desk tickets with intelligent pay stub assistants, Hubert shares real-world insights on balancing automation with the irreplaceable human touch in HR. Learn how Quest is consolidating 95% of their HR platform into Oracle, achieving measurable cost savings while preparing managers to work smarter through AI-powered tools. Whether you're curious about navigating data governance in healthcare, measuring AI success metrics, or understanding where human judgment still plays a critical role, this conversation delivers practical wisdom from the front lines of enterprise AI transformation. Tune in to hear why the future of work isn't about replacing humans, but about augmenting individuals to focus on what truly matters.
Stream it now, and be sure to LIKE, SHARE, and COMMENT!
​
Transcript:​
00:02
Greetings all. Welcome to the enterprise edge podcast. We are recording from Oracle AI world here at the Venetian in Las Vegas. And I am joined by Hubert winter, senior director of HR, HR Technology at Quest Diagnostics. We have a packed agenda here. We're going to we're going to, we're going to go through some questions with Hubert, understand his history with the company and his use of Oracle past, present and future. Before we do that, let's give Hubert a proper welcome. Hubert, thanks for being with us. Oh, absolutely great. Thanks for having me. You're welcome. And Hubert lives in Brooklyn, New York, originally, originally from Germany.
00:40
So he is a, he is an eclectic gentleman. We're happy to have him.
00:45
So we're here in Vegas. We're an AI world. Maybe we'll start with a kind of an opening warm up. Have you seen anything? Heard anything, observed, anything here in in Las Vegas, that has piqued your interest, as far as the Oracle ecosystem and what they're bringing to market, the Oracle ecosystem, I thought the gambling downstairs,
01:10
all the money, exactly, the casinos.
01:13
AI, no, but, I mean, we are part of the soap testing for Oracle so we get our release a bit earlier to do our own regression testing. So we have now 20 5d available, the latest release.
01:29
And so as soon as it came out, I wanted to have a full installation of the AI agents again, and found that there are many more now that are coming delivered. So one is called manager concierge.
01:45
And the really cool thing is, so what we want to do is roll out the team activity center, that's the successor of my team in HR, and then have as a guided journey the manager concierge. So the manager can then ask questions like, How long was John on vacation this year? What can I do with the compensation of chain, or whatever it is? And the manager can ask these questions, and we want to try that out, starting late in the year, when we will get, officially, 20 5d at least as an HR pilot to get more information. Because I think one of the things with AI and the answers that it provides is you have to be a bit careful, and you need to have a pilot first before you roll it out to many more people. I just think, because now this kind of I ask a question get an answer, is very, very new. I mean, we all may use it as part when we because no one is searching anymore. I think in Google, everyone is now asking questions and expect the right answers. But because of that, the change management shouldn't be so high Yeah, before we have we'll have a lot of people who will like the old style documents where everything is described how to do things. You will have them for quite some time, but I hope that over the next years, this will just go away, and it will be much more intuitive with an agent that can help you doing your work right? As an employee or manager, right? So you so Huber, you're in, you're in an interesting space. HR, even Capital Management. HCM, however you want to call it, right, but it's, there's, there's this interesting, I guess you could call it a paradox, right? You have the need to sort of train and upskill and develop human employees, and you have, you know, specifically on AI capabilities. In many cases, you're training humans in their various roles how to better leverage AI to do their jobs, right, whether those are generative tools or agents that are assisting them, or whatever, right? But then you also have the sort of the remit, in my view, to leverage AI, to augment the human roles, right? There was a who was it? Was it? The guy from Avis Budget this morning. I like what he said. He said that AI should stand for augmenting individuals, and not artificial intelligence, because you're augmenting individuals, right? I like that. It sort of dispels a lot of the fear people have anyway. I don't know if it's a paradox, but you have this idea of you're investing in humans, and you're now in also investing in agents, all of which are now part of the workforce. You could say, I mean, how do you decide, and either individually or with your partners at quest which skills are going to be owned by humans and which ones are going to be driven and owned by agents and robots. Is that too simplistic? Or maybe a bit okay? Because on the one hand, you also have a lot of people who need to reskill, so you create a new kind of engineer who needs to write prompts. I mean, you have to build that, but you're absolutely right. But, I mean, we do that since many, many years to automate. Yes, and this is some kind of automation that will come in and if, let's say managers are more efficient in their job and employees are more efficient in their job, then it helps everyone, because then cost will go down dramatically. So in so far, it's another kind of automation that we definitely need. So I wouldn't see it as Yes, I mean, you will have efficiencies driven by cost reductions and where they are coming from. That's something we have to figure out. Got it? I like that. That makes sense.
06:01
So let's talk about a factor that it's, it's under the umbrella of trust, right? So that there's, there's a lot of discussion here and everywhere I go around trust. And you to your point about piloting and testing and making sure that everything is in compliance with whatever regulations or standard procedures or what have you, right? And so a lot of what Oracle is bringing to market, and a lot of the enterprise application providers are bringing to market, requires an element of trust. It requires, in my opinion, an element you have to trust. Certainly have to trust Oracle. You have to trust the provider. You have to trust the intelligence that's built into the solution. What data is the model trained on. Larry was talking fascinating a moment ago about public data and private data and how you can sort of knit it all together for the sort of like optimal source of data for, you know, precise tasks right in certain industries and certain geographies and whatnot. So I guess the question underneath all of that is,
07:10
are you finding that health care workers, because Quest Diagnostics is is sort of in the health care space, generally speaking, right? Are you finding that given the sensitivity of a lot of the data that that I think I'm not in this industry every day, but I think there's a lot of sensitive information, patient data, identifying information, privacy has always been an issue, always been a concern, already, and now perhaps it's even more of an issue now that you're using it to train AI agents. So I'm wondering again, you keep me honest, is, is the is the AI era elevating the importance of trust and security and privacy in the healthcare space that quest operates in or is it just standard, business as usual? You've dealt with these issues, and this is just another wrinkle of the same story. I think because we have the experience, whether it's patient data or employee data, we know how to handle it. We have the right governance tools in the company. We have an a governance board that will ask the right questions. So it's nothing new to us, but it's definitely different. Yeah, that is definitely correct. Yeah, it's not something that we have had before. We have to ensure that none of our data goes anywhere outside, because that that would be just wrong. It's the same rule.
08:43
There's no difference then before, and it's just new technology and but if you apply, so to say, the old rules to the new technology that no date that the data is fully protected, then I don't think we have this issue. Yeah, yeah, but definitely something to look out for, yeah, and are you subject also to
09:06
data residence regulations like sovereignty and things like that, around the world, like, where you the data? Most of our employees are in the US. In the US, we have a few in Mexico, Canada, even in Germany, we have 18 people.
09:23
Do not know them all by name, yes,
09:27
and but as for us, the center, the focus center, is really the US at the moment, got it, got it, got it so that we don't go out with that and got it, yeah, no, I'm interested you were in measurement and how, as you guys adopt and deploy and you're consolidating to Oracle, and how you, how you measure success and satisfaction with the outcomes that you're seeing. Right? You're, you're getting early access to solutions, and you're testing those, and you're, you're, obviously, you've got, you've got an opportunity to maybe optimize as you're testing those. And I'm curious maybe, maybe some examples of of actual metrics or measurement techniques that you're using to assess whether or not real value is being created, whether that's efficiency, or whether that's customer experience or some other sort of tangible value. What kinds of maybe a couple examples of those metrics as a one is definitely since we consolidated most of our HR platform, I would say 95% into Oracle. Is cost. Cost that is definitely something. From a less need for developers to create complex integrations to asking users for questions around where to go. That's
09:57
10:54
definitely something.
10:57
The second point is also the number of HR help desk tickets that we have where we can really measure what kind of questions do people have.
11:09
I mean, here, the issue is, as soon as we see certain trends, we use these trends and then say we have to do something in that space. So we used, for example, the pay stub agent that we introduced just to explain people more about how this pay stub works, what kind of deductions do you find as it is a rack agent, what kind of deductions you find on it? What do these different terms mean?
11:40
And with that, you see a reduction in the number of tickets. Now, the only issue is it takes time, because people like still to create a ticket or to call or to ask a question, to ask an agent a question, a non person, aggression is something new, and we have to get people more and more used to do that. And I think they will like it, because the answer is better than asking someone and waiting to call and all that. So in so far, that's another point. I think we will see a reduction, which we already saw as a reduction in tickets, and maybe last one is but this is more prognosis of me, so we don't have that yet. I think as soon as we introduce these combined better as a team activity center, where manager has everything in one place, plus help via agents that can also help managers. So when we do our next employee survey, or I have a customer focus group and I ask them, they already said, Yeah, we like that. Or generating, as one of the great examples is no one likes to write performance reviews or creating goals. And the quality of goals or the quality of a performance review, is so much better with AI than humans spending hours to create them. Yes, yes, because you can create in a couple of sentences. You give what you want the AI to describe, and it says it much better than you be able to do. That's right, that's right in a fraction of the time. Yeah, exactly, exactly. The
13:36
other, the other question I have, and it's a segment, sort of a segue from what we just said there, in terms of the performance, performance review use case, where there's sort of an obvious case where, with a little bit of human prompting, the outcome can pretty much be optimized by by the by the agent or the autonomous workflow, right? So there's an element of human judgment that I wonder in the field of HR, especially given that's your domain, that will always be relevant. There will always be this need for some human judgment, perhaps some of those gray areas that that happen in HR.
14:22
Now you're focused on HR tech. Now HR ops and sort of the operations of HR, you know, handle a lot of very in some cases, sensitive issues and challenging issues, legal issues, sensitive things around payroll and salary and equity and all kinds of things, right? So where do you think human judgment, the role of human judgment in HR over the next five years, is still going to be pretty critical, right? Absolutely, any employee relations cases will be there.
14:59
We will have cases where we just have to make exceptions for employees that, and this will be there, so there will be no automation. Let's say, like, if you come in five days late, you will be let go or anything, right, right? There will always be a human factor, so you cannot automate that, right? Right?
15:21
So I do believe that we can concentrate more on those cases and actually be more human, so managers can have more time with their employees if we give them the time by, let's say, having more help with creating a performance review, or pointing out that these goals have not been met yet, or stuff like this, right? Tasks where they would have to go in the system and check but helping them doing this task and automating it, right? Yeah, and then giving them more time for the special cases. Yes, yes. We need humans, that's right, and not Romans. It's a very good case. In fact, it's, it's aligned pretty well with the examples this morning from from Marriott, right then the same idea, if you check into a hotel, all that punching of the keyboard that they do while you're standing there, a lot of that can just be automated, and then that person could spend a lot more time optimizing the customer experience, working with the person right, and being face to face, and, you know, making sure they're delighted and whatnot, instead of being heads down. Pound, pound, pound, right, similar right to what we're talking about. But again, maybe it's only the special cases. Yeah, because now with many hotels, you can check in totally without that's true. Yeah, that's right, that's right. That's only if you have an issue, then you need suddenly more time. That's right. And if you are the customer behind who wants to just check in, right, right? Cannot, because there is no way for you. You have to wait and maybe even listen to that story, right, that you're not actually interested in That's right, that's right, that's right, that's fair enough. Fair enough. And I think that's the same for managers, yeah. So if you want to just have a punch of a time clock approved as an employee, you don't need to wait for another employee, which needed much more time, right? Right? Your approval for that. All right, they're the outliers. Well, we have a couple minutes left.
17:26
I have a few questions that are they're called Speed Round questions. They're a little more fun. They're a little lighter. We'll dismount. Get a little insight into your your what drives you, what your psyche is, what's the, what's the weirdest thing that someone at Quest has asked HR tech to solve?
17:49
I have backups if you don't want to. Actually, no idea that I could come up. Okay, okay, that's okay. That's okay.
18:03
If you could clone yourself, if we could make an agent out of you to become a battery, like in matrix, that's right, that's that's a little scary, but to handle one part of your job, what part of your job would you delegate first to this agent, how to become better in organizing myself towards people so that I know what they want and how to best react to them.
18:37
So this kind of playing different roles, because it takes a lot of energy. Yes, it does anything political, yes, which is not tech related. That's right kind of stuff where
18:50
most people will say, you always have to be a human for it. Got it? Got it. All right, we'll end with this one. If you were a professor, what would you teach? I have a PhD in physics. Do you okay? Yes, okay, I taught physics. You did teach physics, yeah. So would you do that again? Or would you, would you choose a different route? You would do that again? Well, that's That's great. Well, awesome. Well, it is. It is top of the hour. Hubert, thank you so much for joining us on the Enterprise edge, thank you all for listening and joining us here virtually in Las Vegas. We will catch you on the next episode of the enterprise edge podcast. Thank you. Goodbye.
