Oracle Cloud HCM s New Generative AI Features With Guy Waterman

William Tincup: [00:00:00] This is William Tincup, and you are listening to the Recruiting Daily podcast. Today we have Guy on from Oracle, and our topic today is Oracle Cloud HCM’s New Generative AI Features, a topic that a lot of people, it’s on everyone’s mind, quite frankly, everyone I talk to. Can’t wait to hear what’s going on at Oracle, what they’ve got going, what they’re cooking, and what they’re looking forward to. Guy, would you do us a favor and introduce yourself and what you do at Oracle?

Guy Waterman: Absolutely. [00:01:00] Thanks, William. I’ll tell you the, interesting thing from my perspective is that I’m one of the people that gets to work with all of the really cool technologies and what a great time to be here with generative AI, with AI in general, working with this, my role is to really sponsor all of the great innovations that we’re coming up with from cloud HCM, from our I Oracle Cloud Infrastructure, and surfacing that right into the application space, so it’s super simple for our customers to be able to take advantage of and use.

As the strategist that’s responsible for not only our analytics, for common technology capabilities like integration, Business process management. In addition, I’m also working with our artificial intelligence team and our generative AI team as well. Get an opportunity to dabble in a lot of different areas and hopefully we’ll be able to share with some of the listeners how that comes to [00:02:00] provide benefit for overall our overall applications customers.

William Tincup: That is guy, that doesn’t sound like a job as much as just a lot of fun. I don’t I’m sure there’s a lot of work involved, but that just sounds really fun.

Guy Waterman: It’s one of those things, be careful what you ask for, you just might get it. I feel what you’re saying, and I’m very fortunate to be not only in a great position, but really at a company that tends to be viewed as something that is we’re emerging.

Rapidly within that cloud space and being able to apply the capabilities that we’re innovating. But being able to apply that to an HR and a talent acquisition space is definitely something that provides a lot of excitement in my world.

William Tincup: So I’m assuming that you y’all have been working on AI and generative AI far longer than your clients have probably been asking for it, right?

So y’all have been in the lab, cooking up different things, and then [00:03:00] ChatGP TV, it’s launched, OpenAI, people start to play around with it a little bit more, and then maybe more practitioners raise their hand and go, Hey, how does this apply to what we do?

Guy Waterman: Am I right about that? Yes, because when you think about AI, AI has its roots back in operations research.

It’s a lot of really, we can go back and take a look from remote, the robotic process automation craze that we were really emerging from three years ago. We can go beyond that one and look at some OR type capabilities and really dig in there. But from our work within the cloud, the first recollection I have of actually AI surfacing up was with our deployment of advanced controls within the application for both ERP.

And for HCM and to see how that has really helped us get our feet wet from introducing customers to what is this thing called artificial intelligence [00:04:00] and what could it mean to you? And then gaining their trust has been that’s always been the thing that I look at is that everything AI sounds a little bit too auto magic for me.

How can I actually prove to you all that this is something that can help deliver value? Reduce risk and make life better for your workforce.

William Tincup: So let’s go through some of the features because I want to get back to the prove and training and getting people to understand. But before I do that let’s talk about some of the things that you’ve already started to roll out with the Oracle’s Cloud HCM.

Guy Waterman: Oh, absolutely. Yeah, the from an AI perspective, so artificial intelligence, it was once called adaptive intelligence within within our capabilities, because we were branding very specific functions where we knew that AI could impact the lives of your recruiters, the lives of basically people that are considering to coming to your organization.

And AI for us, from an HCM standpoint, started with that recruiting cloud [00:05:00] capability for looking at the next best candidate, allowing the candidate to look at the next best job. Perhaps they’re looking at an opportunity and it wasn’t exactly the right fit, but what are the right fit for me? And then being able to look at how can we reduce that time to hire?

How can we not only compliment the fact that we’re speeding the type of the, or the acquisition of the candidate, And the process, but we’re also looking and identifying the best fit candidates. So we may be able to hire fast, but once we got them here, we realized they weren’t the right person. So we took that approach of saying, how do we not only make sure that we’re being able to do things where we have an adequate supply of resources available?

But then we’re also looking at, are they the right people culturally? Are they the right people from a local position? Do they have the right history? And based upon the trends that we’ve been able to recognize, that they are going to be the best fit for your organization. And that’s where we started from an [00:06:00] AI standpoint.

What we’re seeing now is that this leads into skills and skills ontologies, but most recently is really one of those things that we talked about with that whole generative AI that you mentioned just a little bit earlier.

William Tincup: So the first one, I’m glad that you broke that down because a lot of people were trying, especially with matching when it first came out is let’s speed everything up, which is great.

Our candidates want to go faster, hiring teams want to go faster. So speed’s great. But I think we’re to a phase now where speed is still important, but quality. is, if not just as important, might be more important. And so you’ve now that qualitative layer where you’re, again, as you said, it’s chemistry and potentiality and skills and the desires of the candidate, desires of the team, all the, all kinds of different things, ways of looking at it.

How did that tell us how that came to be? Is that is that something that the recruiting folks, [00:07:00] your recruiting team and and the folks that work on the recruiting product kind of said, Hey, this is what we need, this is what we’re seeing, or is it, I’m always curious about how things get developed.

Is it a customer that kind of raises their hand and says, Hey, we have this problem, or is it someone internal, or is it a cadre of those things?

Guy Waterman: I will say it’s a cadre of those things because it’s that the mix is really where things become different. And within what we’ve done from the recruiting perspective, obviously we have A group of what I would call, selfishly, the most talented group of talent acquisition developers that are around.

And they’ve got a very strong history of working in the industry. But then, the other thing that we’ve noticed is that we also are working in a in an environment. In a marketplace where our customers are extremely talented, and they know what they want, and they understand innovation and what’s possible, and they bring to us the ideas, and it’s a blending of the so the [00:08:00] cadre comment that you made William, has ended up being, we end up having the skilled technologists coming together with the people with a particular need, And then they’re looking for common outcomes.

And in this situation, when we started looking at blending in just artificial intelligence, trending of information, things that are likely to occur, and bringing that to the fore with our analysis of seemingly very large talent pools and trying to resolve very specific issues around time to hire.

around quality of hire, around even finding alternates because we have great resources. That is something that we’ve been able to harness. And then, of course, get the validation from people that are in practice using these capabilities every day to prove that they’re working, to address potentially even some of the issues that are coming around, from applying automation to something that is as personal of where am I going to work, where am I going to be happiest.

William Tincup: It’s interesting because that applies both pre [00:09:00] hire for candidates, but also same logic kind of applies with employees and internal mobility.

Guy Waterman: I love that you went there, absolutely, because that is the logical extension of where is my best next hire? And the reality we jokingly not maybe jokingly, but we say, look left, look right, because the chances are that best resources probably And it is someone that’s a known quantity, and in fact, what we’ve been able to find by applying AI and being able to do the follow up behind the recommendations and the matching that’s made, that we’re finding that those people that are vetted, that are existing and known quantities tend to be better hires and better resources because they have the More known quantities about them.

So there are gonna be less variability associated with those internal hires, whether they’re transfers or they are people that are growing into a position. [00:10:00] And that has led us into areas of how do I make myself known as someone that is interested in this particular career path? How do I help improve my value to the company?

And one of those things that we’ve led into was not only the career pathing, Aspect of what. Artificial intelligence has enabled, but then we’re also looking at how do I also help improve those what are seemingly mundane tasks or things that may not be the most enjoyable part of my job is describing how well I did today or describing a goal and how that goal actually is going to help me achieve my overall business objectives for my annual objectives.

Those are all things that we have found that are part of the same. Process and that AI is really helping coordinate our efforts in a, in an effort to make sure that we have the right people in the right place that [00:11:00] are providing the right level of service to ultimately our customers.

William Tincup: It’s funny as it, it reminds me of JIT, just in time things I learned in business school more on the manufacturing and operations side back then, but we’re talking about talent in much the same way.

Thank you.

Guy Waterman: Interesting that you would bring that up because when you’re talking about JIT and we get back into just in time and you start looking at supply chain type capabilities and replenishment capabilities and then the whole order management, it is a balance of supply and demand balancing.

It’s how can I use known quantities to achieve the appropriate balance? And those balance can have many different variables, but it is very similar. And you’ll find that when we start talking about a lot of the three letter acronyms, things that come out there around the different technologies that are capable, what we’re really getting to is this age old business [00:12:00] problem of balancing supply with the right kind of demand and then getting the appropriate resources in place so that we can come up with the optimal output.

William Tincup: We should probably back up for the audience, because you and I just geeked out, and Sorry. No. We probably should delineate for folks the difference between AI and generative AI, so that they, if they don’t already have a working idea of what that is let’s give them an idea of how you look at that.

Guy Waterman: Absolutely. So AI, and then it’s a great question and a very valid question and don’t ever have anyone feel like they’re asking a question that they should already know the answer to because the reality is. The definition continues to change when we, it’s basically it’s yes, I kid you not William, because as we look at this I know that coming into this role, I absolutely had an appreciation for what was going on within the world of AI and looking at what was possible, [00:13:00] but then into the role where that I’m currently in from an AI perspective, six weeks into it, It was a student body, right?

The definition changed, and then we came back and it changed again. So it’s always worthwhile to clarify with an organization what do you really mean by AI? And AI is really nothing other than saying, we, there are a lot of knowns, and we’re trying to help them predict or come up with the most likely trend that’s going to be the outcome.

And then in between that. is the ability for us to influence the way that those decisions are made, and that’s the one that always gets people a little bit nervous, because, yeah, anything from a mathematics standpoint, if we know to put the right kinds of inputs, we can influence the outcomes, but the Understanding that and what the risks are from an AI perspective, I think that we’ve all been comfortable enough with forecasts, we’ve [00:14:00] been comfortable enough with, ordering enough material so that we have enough of these products on the shelf for our customers when they come through the front door, that we’re comfortable with that, but when we start applying it to something that could potentially be a manipulated outcome People begin to get a little bit nervous.

So from an AI standpoint, we’re really just trying to say, what are the likely outcomes based upon The things that are known. Now that’s different than what we’re talking about from a generative AI, because in general, when we’re talking about generative AI, this is going to be something that is more of a narrative.

It’s not just an answer or a specific recommendation. It becomes an ex an explanation. It might have some dependencies associated with it. And by the way, I’m referring to this from a, an assisted authoring type capability, where we’re getting suggestions or even a narrative associated with something or maybe even summarization.

Hey, could, [00:15:00] someone just spent a half an hour telling you about themselves on a job application. Can I get something that goes over the top of that and summarizes the best three sentences That describes who Guy Waterman is. That’s really what we’re looking at from a generative AI story, is that we’re trying to tell a story, and then what we’re seeing is that having those capabilities frees people up so that they can even become more creative and then focus their resources, their talents, their efforts in an area that’s ultimately going to help show that they are the better off.

Candidate for the job, that they’re a better fit for the job that they may actually have other talents that may be un Benonce to them, but maybe of value to your organization as well.

William Tincup: Two things that I usually, when people ask me that question or something similar, I say, listen, with generative AI, write in pencil.

That’s because, don’t carve things in stone, because, and don’t rewrite things in pen. Just keep it in pencil [00:16:00] for, the next decade or so, and then maybe in our lifetimes, we’ll get to a place where we actually understand what it is. The other is, I talk about. Calibration and recalibration with both AI.

It’s like understanding that these are algorithms and algorithms there’s guardrails to these algorithms and they don’t always do exactly what we want them to. There are some, sometimes when they veer off a little bit, which is fine. It’s understanding that’s a calibration issue. Like we have to understand that this needs to be calibrated slash recalibrated and I, that it helps.

I know that I know it’s helped people when I’ve talked about it like that. It’s listen, first of all, I get it. The ground beneath you is moving as you’re walking across it. I understand. I understand how unpleasant. Maybe that might be or might seem, but think about the benefits, think about what we just talked about in hiring and how you can get, both on the front end with pre, pre hire you can get candidates faster and at a higher quality level, [00:17:00] and you’re going to save a bunch of time that was wasted doing other things that you just shouldn’t do.

Like once you see that. Okay. You know what? I need to get comfortable with being uncomfortable. Yeah. Okay. You can see it sometimes even in conferences. You can just, you can see that once. They’re over that kind of emotional hurdle. They’re like okay. You know what? I get it. I get it.

It’s like the beginning of the internet for those of us around long enough to remember the beginning of the internet, it was weird for a while. Like people don’t remember that now things are, I say normalized but back then, the late nineties it was not normal. Nothing was normal.

And so I did want to ask you. A lot of people have talked about generative AI and in the sense of like it being a passenger or a co pilot of sorts.

Guy Waterman: Do you see it that way? Not to use a product name or anything, right?

William Tincup: Good point. Good point. Good point.[00:18:00] I watched Dexter a long time ago and they called it the Dark Passenger.

So I try not to go to say Dark Passenger, but the kind of driver passenger model where there’s something around you and it’s helping you solve problems. Yeah. And like you said, take this 3000 word document. Give me the three paragraphs that I actually need to pay attention to that, that’s something, first of all, that’s super helpful, but also you’ve got to be able to ask it the right questions.

Guy Waterman: Two things on that one to respond to your answer and then add on to that last point, if I may one, every day in the last, I get multiple emails and phone calls and communications coming from many different sources saying what is, what does generative AI mean to me, and what does it mean to Oracle, and what does it mean to my future?

So the uncertainty is definitely rampant. I [00:19:00] do believe. William, that people are believe, people do see the potential for improvement and for the ability to do good. With this capability, with these technologies, with the assistance that this can provide people understand the value. It’s inherent, and it has been so long since the appearance of value or the understanding of what this represents to us has really been that apparent, and I go back to probably 20 years ago the last time that I saw this level of excitement was around the time when Wireless internet became generally available, even with the old PC MCIA cards that were sliding into credit card size cards, but the people were talking about, can you imagine what it’s like to always be on with not having to have to dial up and all of that?

Can you imagine how great that’s? And some people couldn’t, but now I have [00:20:00] a generation of workers that. That’s all they’ve known. And how is their life different? And that’s what’s become, I think, different from be aware that the world beneath your feet are shifting. The world is shifting. The ground is shifting.

And that people are shifting. are actually uncomfortable with the fact that’s happening. I liken it back to that 20 years ago when we saw people were uncomfortable with this always being on and what are you going to monitor to today people that actually embrace the always on culture and the fact that there are things that we never anticipated with streaming entertainment or with the ability to come in and start up a meeting when I’m you know driving down the road and I’m all I’m connecting with is a with, as a mobile phone, and I’m getting video, and I’m getting voice, and I’m getting answers to questions that I need, and I think that’s what’s happening, and that’s what we’re seeing with generative AI, is that people don’t necessarily have to look at this and say, [00:21:00] this is a risk to my particular job or to my well being, because what you’re doing is freeing me up.

The mundane tasks, things that I granted you might be able to, as a generative AI engine, come up with a more descriptive and accurate job description or summary of what we’re looking at, or even better, recommendations. But the reality is, what I’ve tied around those things, what I’ve done to connect that description to this individual, to this opportunity, That’s what’s innately human.

And there’s nothing artificial about that. We all are going to be needed to be able to make those connections that up to this point, we’re not able to do from an automated standpoint. I think that will represent a tremendous amount of value. And quite honestly, the next generation will be looking at us saying, oh, we, we remember when people worried about that.

William Tincup: I remember being worried about that. I know you wanted to talk or it seemed like you wanted when I talked about asking questions or prompts. Oh, yeah. Yes. Do we, do you feel [00:22:00] like especially, being, Being Oracle, do you feel like we have you have a responsibility with your customers or prospects even that kind of teach them how to ask questions or how to ask

Guy Waterman: prompts?

And so yes, and we also have a. A large number of product experts that know the solution that will do that for you, because, and this has been, and maybe it’s just where we are in the maturity of the market at this point, William, is that I’m, and that’s what I think it is, because as we all learn about prompts, and by the way, prompts is really nothing other than gathering.

Information that I can provide to a generative AI engine, or to a a solver, basically, something that’s going to provide an answer, And expect a specific response. But if a prompt is basically guiding the kinds of questions so that I [00:23:00] can get a relevant response, right? And that’s what we’re referring to with prompts.

And we are engineering prompting. We actually have the process. It’s called engineered prompting within our applications. And the goal at this point in our, and I’ll say it’s admitted infancy of working with generative AI, albeit extremely powerful. Our goal is to just give The user an opportunity to select whether they want to get generated content or they don’t want generated content and without having to do anything other than tick a button.

And get the response. And no harm. If you don’t want it, that’s fine. You can go ahead and choose to ignore it. Or you can regenerate it. Or you can keep some of it, modify it, save it. But the goal here is that the prompting needs to be pretty quick. And it needs to be intuitive. And I use this as an example with some folks, is that mobile applications, and I’m a bit of a [00:24:00] mobile person from way back when but this one always blew me away, and this is what’s at risk with generative AI, and with talent acquisition, and particularly with people that are in these high stress type positions, making decisions about who you’re going to hire, and are they a good hire, but how long does the average individual retire?

Give a mobile application, once downloaded, to work, and it’s less than five seconds, and how many times do they not even go back if it doesn’t work? I really appreciate you taking on this topic, because this is a critical… And I think that’s the success of generative AI within enterprise applications, is that we all have one shot to make this work, one shot to mess this up, if you wanted to look at the counterpart part of that, but we have one opportunity to prove out that this is of value to you, and that, hey, this could actually work, and that I am willing to continue to [00:25:00] invest in this.

And that’s where we are with generative AI at this point. It is a critical point. It is so early, and it’s strange to say that it’s critical yet, and it’s I’m still as young as it is but there is a lot of value that exists here. And I believe that people look at it and say I like what I’m seeing.

It’s easy enough for me to use. I’m willing to go one step further. And that’s exactly where we’re focused with recruiting cloud capabilities, with harnessing the right kind of technologies with making sure that the response times are there and to make sure that you’re not getting crazy responses to what you thought was a pretty logical question.

William Tincup: It’s interesting. I talked to somebody earlier this week and they said, listen, we just did a kind of a small, a report. And basically one of the things we’ve learned is about 80% of the practitioners we surveyed are using some form, probably open AI or something like that. They’re using some form of of popularized AI and 80% of them are terrified.

[00:26:00] So on one level, I think, and it mirrors everything that you’ve said, it’s there’s excitement at the same time that there’s this huge excitement. People see the world in a different way. There also, there’s an anxiety that’s there as well. So it’s mitigating that anxiety, teaching, training, helping people understand, Hey, it might not be perfect the first time, et cetera.

Let’s train you. Make sure you get that. That answer that you want, that’s tailored to, exactly what you’re seeking out. You might not get there first. Like when I first started talking about artificial intelligence, I’m like we can call it artificial, but we shouldn’t call it intelligent because none of this stuff’s intelligent to start with.

It will get there. It’ll get there fast and it’ll get there over time, but let’s, none of it’s intelligent. But, thank you. Thank you. Guy, we’ve got to talk about every three months because I just want to window into what you’re doing and what you’re seeing. And thank you so much for coming on the

Guy Waterman: podcast.

Hey, I appreciate it, and I’ll tell you what, in the next three months, I can tell you what’s next on the horizon as well, because I think that we looked at these [00:27:00] things, and today we talked a lot about text generation, we talked about a lot of content generation. Hopefully, if you’ll have me back, we’ll give you an opportunity to see really what’s coming and what is truly possible, and we’ll have some tangible results to share with you at that point.

William Tincup: Awesome. Awesome. Thanks. Have a great day, and thanks for everyone listening. Until next time.