Ep 137: Trading AI Hype for Evidence and Education in Talent Tech with Hayley Skivington

[00:00:09] [SPEAKER_00] Hey everyone, it's Bob. Welcome back to Elevate Your AIQ, your go-to source for insightful conversations on human-centric AI readiness, talent transformation, responsible innovation, and the future of work. This week, I'm joined by Hayley Skivington. She's the head of product at Olio. Started her career in recruitment and has come full circle to shape the technology recruiters rely on today.

[00:00:30] [SPEAKER_00] We talk about why vendors now need to act as educators and trusted advisors, how Olio competes in a crowded market by complimenting the systems customers already have, why showing recruiters and candidates how AI reaches its conclusions is key to rebuilding trust on both sides of the hiring table, and how hands-on AI literacy, starting with something as simple as planning your family's weekly meals, takes much of the fear out of the technology.

[00:00:56] [SPEAKER_00] Hayley also shares how one public sector team went from banning AI outright to embracing it across their entire organization, and why she believes the human touch in recruiting remains irreplaceable. Tune in for a candid and practical conversation, and thank you, as always, for listening to the show. Let's go talk to Hayley. Hey everyone, welcome back to another episode of Elevate Your AIQ. I am your host, Bob Pulver, and with me today, I have Hayley Skivington. How are you today, Hayley?

[00:01:22] [SPEAKER_03] Hayley Skivington. I'm good, thanks Bob. And you?

[00:01:23] [SPEAKER_01] I'm doing very well. I'm excited for this conversation.

[00:01:26] [SPEAKER_03] Me too, me too. Long time in the planning.

[00:01:29] [SPEAKER_01] Yeah, you and I have spoken a couple times about some topics that we will definitely cover around responsible AI and ethical use and ethical design, and a bunch of other things you're working on in the talent acquisition space. So, Hayley, why don't we just start with just you giving my listeners a little bit of your backstory and your career path through product roles and go-to-market roles and things like that in the HR space.

[00:01:53] [SPEAKER_03] Yeah, it's quite interesting. And when we have these conversations with people about, oh, you're a head of product, but you've come from go-to-market and product marketing. How does that work?

[00:02:32] [SPEAKER_03] And how can I help them to overcome their challenges, but make it as easy as possible to do so? So, yeah, bit of a journey. I've started a long, long time ago. I started my career out of the gate from school in recruitment and now full circle kind of quite a few years on. I've ended up kind of had a product in the talent acquisition space. So really exciting, really exciting time at the moment as well.

[00:02:57] [SPEAKER_03] You know, we've kind of gone from ATSs being where they are and we're looking at can we expand or do we kind of explore diversification to this huge explosion of AI just generally, but particularly in the talent acquisition space. So, you know, kind of taking my background and I said, OK, how do we overcome those challenges?

[00:03:17] [SPEAKER_03] Not just in the technology sense, but actually when we think about the recruiters and what AI kind of means, you know, there's a lot of fear still with AI in terms of your own job or litigation or what happens if I get it wrong. So, you know, I feel like actually the role as head of product at Olio has kind of expanded into that. We need to hold our customers' hands and our prospects' hands and be trusted advisors.

[00:03:42] [SPEAKER_03] And, you know, we need to be at the front of this technology so we can educate and share that knowledge as well as designing products to meet the modern challenges in recruitment. So, yeah, a little bit about me and how I approach product design.

[00:03:57] [SPEAKER_01] Yeah, I mean, I think the logic is sound, of course, starting with the actual business problems, you know, where are we, you know, falling short with existing metrics or are we not tracking, you know, the metrics that truly matter and are going to make a difference? And how can we work through that? A lot of involvement with users of which, you know, you I'm sure were one at one point.

[00:04:24] [SPEAKER_01] And because you're using this in recruitment, I would imagine you're drinking your own champagne, as we like to say, right, in terms of, you know, the platform and how it helps, you know, Olio's own talent.

[00:04:38] [SPEAKER_03] Yeah, exactly. And that's how we always start, you know, our head of people is kind of our number one tester for all new innovations. You know, we don't recruit at the scale of some of our organizations, but actually from that perspective of can you use it? Is it solving a problem? That's always where we want to come from, you know. And me and the head of technology are always laughing because whenever anybody comes to us with anything, we always say, what's the problem?

[00:05:07] [SPEAKER_03] What are we trying to fix? What are we, where are we going to deliver that value? It's very easy, particularly with AI to be like, wouldn't it be amazing if or how about we did that? And then she, if it's not, if nobody's got that problem or nobody needs that to overcome something, it's very difficult then to get people to adopt it, to live in the commercial business world.

[00:05:30] [SPEAKER_03] So, you know, I do think utilizing our staff to test out our products does, you know, and she's brutal. I'll tell you, Bob, the feedback is very candid and honest, but that's great because we have a confidence then when we start to take it to customer champions and early adopters before them, full productionization and taking it, take it out to market, you know, it's very much how we work.

[00:05:55] [SPEAKER_03] And our customers definitely influence kind of where we focus our attention and those big meaty problems to solve. They're very open and honest about what they're facing in the market, which is great for us because we've got that insight coming in and we can kind of shape the products to make sure we're overcoming those.

[00:06:13] [SPEAKER_01] Yeah. Now, when you think about the AI sort of injection of, you know, capabilities into, you know, product suites and systems of record, like applicant tracking systems, like how are you approaching it? I know you've spent a lot of time talking about how do we sort of demystify, you know, what we mean by AI,

[00:06:38] [SPEAKER_01] making sure we can overcome people's, you know, fears and doubt, uncertainty. And so how do you approach that? Because I feel like that's just a whole other layer on top of what you're trying to build to make sure that you get the, you get satisfied users, that you're solving those real problems, that you're not over-engineering a solution or being like a me too or, you know.

[00:07:05] [SPEAKER_01] Oh yeah, we now with, now with more AI, you know, those kinds of things. I mean, I think you're going to have a lot of skeptics and it's a crowded market. I mean, you've got a head of product undersells. It's obviously a big job and a big title, but it also in some ways I feel undersells all that goes into doing your job successfully.

[00:07:28] [SPEAKER_03] Yeah, it is. Like at the moment, I think there's just challenges from most directions. And some of what you said is completely accurate. You know, we have to avoid the me too. You know, when we think about what we can do with AI, it's very easy to want to be a kid in a candy shop and just want to check AI at absolutely everything. But the foundations and the core of developing anything is you have to have strong foundations.

[00:07:56] [SPEAKER_03] So, you know, we're kind of learning as we're going a little bit here and kind of saying, you know, we want to improve this area of the product. Are our foundations sure enough? Is the product, you know, infrastructure strong enough and sufficient enough to work well with AI? There is a little bit of a kind of idea that AI is a bit of a sticking plaster, a band-aid over bad process or bad infrastructure. And that actually isn't true.

[00:08:22] [SPEAKER_03] It needs all of that stability to perform in the way that it does. So we very much are kind of looking natively at, okay, how do we bring AI into the ATS in a native way? And what do we want it to do? So workflow automation has been at the core of Olio for 30 years. How do we support recruiters to be more efficient so they can focus on what really matters in recruitment?

[00:08:49] [SPEAKER_03] And I think that's even more true now when we think about AI, like where are our bottlenecks and sticking points? And how do we accelerate those processes? But accelerate it in a way that one, delivers value, but two, enables recruiters to make those decisions and have the information they need to make it with confidence. Because that's where that black box, that gray area comes in.

[00:09:14] [SPEAKER_03] If it just does it and you suddenly get 10 candidates listed for you, why have you got them? What has the system done? Why are they the right candidates? So everything we're looking at is kind of saying, okay, if we do this, what is the AI doing? And how do we crucially show the recruiter what we've done? So when we think about screening as an example, you know, we can go through and we can say,

[00:09:40] [SPEAKER_03] we've gone through thousands of candidates in moments because AI is that quick. But actually, based on what you're looking for, here are the candidates that not only are meeting your criteria, those skills that you're looking for, but this is where we've seen the evidence and this is how the AI has done it. So as a recruiter, you're informed, you understand and say, okay, great. I can see in their resume, they've got four years.

[00:10:06] [SPEAKER_03] But actually in their, you know, when they've been describing a particular challenging project, they've also evidenced these key skills and elevated their experience there. So I understand that candidate. So I feel empowered to make that decision. And that I just think is critical to recruiters to overcome their fear because they can see what's happening. You know, it's not magic.

[00:10:32] [SPEAKER_03] It's just doing what they would do, looking at resumes and bringing it together very quickly for them to then make their decisions. But critically, it gives that information for the recruiter to then go back to a candidate. If a candidate says, well, hang on, whoa, whoa, whoa, I've got that. Okay, this is what you submitted. This doesn't demonstrate what we've asked for. So the candidates can also learn from it because that's where, and Bob, I'm sure you speak about this all the time. There's like a bit of a friction.

[00:11:01] [SPEAKER_03] Recruiters and candidates, they're kind of head to head at the minute. Well, you're using AI. I'm using AI. Nobody's being honest and truthful. So by using our infrastructure to surface that and say, well, here's why, it brings back that trust element as well because people can say, well, look, we're using the tools. All it's doing is highlighting where you're demonstrating what we've asked for. As a candidate, you can learn from that and adapt and improve your next applications.

[00:11:29] [SPEAKER_03] But as a recruiter, you've got that data to make your decisions. Data is always going to be key, whether we're 15 years ago where data transformation was the big conversation to now AI transformation. It all comes down to what you've got available to you and how you can make those decisions.

[00:11:46] [SPEAKER_01] I was just going to say, I mean, you're hitting almost one by one some of the elements, you know, aspects of what we mean when we say responsible AI with a capital R. I mean, transparency, accountability, you know, this recruiter is still on the hook for making a decision. They're still accountable for the decision that's made somewhere, you know, mid-funnel at one of the steps that these candidates are going through.

[00:12:15] [SPEAKER_01] And so it is awkward, right? Like everybody wants this win-win relationship, but we're almost like, you know, manipulating each other along the way. And then it makes you question, you know, have we actually achieved the results that we want? Is this person going to stay? Were they at quality higher? Are they engaged? Are they, you know, living up to the potential that we saw through the process?

[00:12:45] [SPEAKER_01] And so it's awkward. I mean, I didn't grow up in HR. So, you know, maybe it's, there's always been a sort of cat and mouse game in recruiting, but I think it's our job to help people realize that AI is here to help at the end of the day, not to exacerbate existing challenges. But I think it's, for some, it's a mindset shift and it's hard, right?

[00:13:15] [SPEAKER_01] Because the market is difficult right now with all the other AI fears about it actually taking, you know, work from people or disadvantaging, you know, certain people. There's a lot, as I was alluding to at the beginning, I mean, there's a lot for folks like yourself to sort of navigate. And there's a balance between how we use this technology for efficiency as well as effectiveness.

[00:13:42] [SPEAKER_01] There's a lot, a lot to juggle.

[00:13:45] [SPEAKER_03] It's a balance, isn't there, as well? Yeah. Because, you know, you want to, you don't want to hold people back from utilizing new technology. But with something as powerful as AI, I think it's really important that there's a level playing field right from the start. You know, tell candidates what's acceptable, but also tell them what you're doing. So you're clearing, you know, you're clearing out that vagueness and opaqueness in the process right from the start. So then candidates will feel more empowered to be more authentic in what they're doing.

[00:14:15] [SPEAKER_03] You know, but there's a responsibility on organizations around AI literacy of their teams. You know, we need to think about, do people understand, number one, the underlying technology, what artificial intelligence is, and the foundations of how it works? I'm not expecting everybody to understand all of the large language models and kind of everything they do, but the premise of what they're able to do, because that's the scariest part for most people is they don't really understand the technology.

[00:14:45] [SPEAKER_03] You know, my best advice to everybody has been probably over the last 12 to 18 months, use it in your personal life first. So, you know, use a Gen AI to create you a menu for you and your kids for the week or help you plan your next vacation, whatever it might be, because then you understand kind of the to and fro that's needed. You know, it does need a level of instruction to perform how you want it to.

[00:15:12] [SPEAKER_03] It doesn't know you out of the gate. So you kind of have to teach it and get it used to your rhythm and give it guidelines and parameters for it to be able to perform how you want it to. And I think that takes a lot of the fear out of it when you realize how much you can control it. You know, if you just say, I want a meal plan for me and my kids, you can get anything. But if it's, I want a meal plan, my children are this age, one only likes chicken,

[00:15:40] [SPEAKER_03] one won't eat carrot, you know, whatever those funny nuances might be. You know, week one, you'll be tinkering. Week two, it'll be like, oh, I remember that kid only eats chicken. So everything comes through and it's kind of learning from you. And that's why I talk about foundations and data. It needs that information to perform effectively.

[00:16:02] [SPEAKER_01] Yeah, I think the sort of explainability of some of the capabilities is a key part of the literacy around AI. Like how did it, you alluded to it before, Hayley, around like, you know, show me the evidence. Like I'm not seeing it with my human eyes or maybe I'm just not picking up, you know, what was inferred or what was implied in a bullet point on a CV.

[00:16:32] [SPEAKER_01] What's explaining the inference and the strength of that inference. And then to your point around, you know, personal use, it's funny, I was just talking to my parents the other day and they were so impressed and almost humored by the way that I think it was Gem and I was interacting. They were trying to make a schedule for some sport, a tennis or golf or something like that.

[00:16:59] [SPEAKER_01] And everyone had different time preferences and availability and vacations and all of this stuff. And they were just amazed at the fact that it was able to sort of decipher all of that and then come up with something that worked for everybody similar to the meal plan. And I mean, those are the kinds of experimentation that I do think are incredibly valuable because it shows the importance of context. And then, you know, maybe balancing priorities and things like that.

[00:17:28] [SPEAKER_01] But I do think the more you use it, the more you start to understand some of the logic behind it or maybe where it wasn't logical because it missed context or forgot something or whatever. I mean, these things happen. They're not, you know, infallible. They're not calculators. No, exactly.

[00:17:49] [SPEAKER_03] And that's a really good point. They are not calculators. That's why that human over the top is always going to be critical, particularly in recruitment. You know, they're not infallible. They can make mistakes. And, well, they can infer something incorrectly. I've, you know, I've seen it where, you know, you've put something in and maybe you've phrased something a little bit oddly. And, you know, normally me, I'm distracted by my kids.

[00:18:15] [SPEAKER_03] And so I've kind of gone back halfway through a sentence and just, yeah, yeah, yeah, that'll be fine. It'll understand what I mean. And it's come back and I've gone, oh, that's not definitely not what I meant. You know, so you have to go back through and say, oh, no, I didn't mean this. Can you reply? You know, but you kind of learn by making those mistakes and doing it on those kind of more trivial tasks. I think it kind of, it sinks in a little bit more and takes some of that fear away.

[00:18:42] [SPEAKER_03] Because I think as well with AI and talent acquisition, it's quite interesting because not only are we trying to overcome challenges and we've kind of got this fear and we've got this need to understand the technology. Then we've got all of these regulations that are changing and emerging and shifting. And, you know, they're taking a while to bed down. And I know, you know, there's new state regulations coming out.

[00:19:06] [SPEAKER_03] Feels like most weeks at the moment, but, you know, kind of, and they're all slightly nuanced and slightly changed. And then we have the EUAI Act plus we have the UK DUIA side. So it's kind of, you know, how do you incorporate all of that? Because they're all slightly different.

[00:19:23] [SPEAKER_03] They're all about mitigating bias and making sure that we don't, you know, create challenges or make decisions against people that's going to directly impact their ability to earn all of those kind of, you know, great things and parameters that we need. But I think all of that unknown around the regulations is kind of feeding the fear a little bit for people. They kind of feel like, I don't know, what am I meant to do? How am I meant to do this?

[00:19:52] [SPEAKER_03] So it's, you know, I think when you're looking at vendors, there's so many vendors in the market and they can all do something with AI. But it's understanding how do you assess those vendors and how do you make sure they can give you what you need to be, one, compliant, but two, make your decisions effectively. And I think that, you know, people are almost hesitant to make a vendor choice because they don't really understand what they should ask to assess them. Does that make sense?

[00:20:19] [SPEAKER_03] We're in a bit of a weird tug of war at the moment about what's a vendor responsibility? What's my responsibility? Which regulation do I need to adhere to? It's really interesting.

[00:20:31] [SPEAKER_01] I feel like trust trumps all other, you know, features, functions, certainly. I mean, to me, if you're a buyer, some of the very first questions you should be asking vendors even to get to a shortlist, let a select a vendor is, do I trust what they're doing? And can they explain how their technology works?

[00:20:58] [SPEAKER_01] Are we clear on, to your point, the sort of responsibility that it's ultimately a shared responsibility if you look at common denominator across some of the existing legislation? I mean, I said, you know, you and I have had conversations with Martin Redstone from Worden in the past. And, you know, Martin and I are a similar find that, you know, we've been sort of shouting this from rooftops for a couple of years now.

[00:21:27] [SPEAKER_01] It's like whenever the EU AI Act goes into effect, this is how it's shaping up. And if you can align yourself to what the EU is doing to protect individuals' lives and livelihoods and personal data, then you'll probably be in very good shape for other legislation or other, you know, geographies, countries in the U.S., with, you know, state-level, even municipality-level, you know, rules.

[00:21:57] [SPEAKER_01] And so you've got to, you've had this view into what is possibly going to happen for quite a while now. So if you're still spinning up shiny objects, you know, we'll fix it later or we'll secure it later or whatever. I just, I don't buy that and I don't like that approach at all. I mean, you should be responsible by design.

[00:22:21] [SPEAKER_01] And so I guess I would hope that more heads of product at your competitors and partners and other stakeholders around the ecosystem, you know, feel the way that you do. I would hope that they do, but I know that they don't all prioritize it the way I would.

[00:22:37] [SPEAKER_03] Yeah, it's interesting. I think some of the nature of our, it's quite interesting because when I go and talk about it, and I remember when I was talking about how we were developing AI 18 months ago, people were a bit like very cautious. And I think it's a lot to do with the heritage of Olio. You know, we work with government authorities, we work with police organizations, we work with financial institutions. So we've always developed in a responsible way.

[00:23:07] [SPEAKER_03] How can we be secure? How can we develop our software in a way that kind of supports these very risk adverse organizations? But being risk adverse doesn't mean they don't want the latest technology. They just need that level of compliance and reassurance around it. So that's exactly kind of how we apply to AI development. Like our processes are in line with ISO 4 to 2001 for good reason.

[00:23:37] [SPEAKER_03] You know, that aligns us, that ensures we have the management systems, the checks, the alerts in place. So that as we're building out, we're building in a very responsible way that we can, you know, validate before we go live. We can put it through rigorous testing. And I think that's critical. I spoke right at the beginning of confidence when we go to market. Confidence for us, you know, head of product, head of technology.

[00:24:04] [SPEAKER_03] We love to know it works and it delivers and we can stand behind it in kind of those real world scenarios before we've gone into market. But actually more critically for organizations, we can give them the information, the data in a very transparent way. So they understand how we've got to where we've got to, to give them a confidence. You know, a lot of the time we're finding, Bob, as well, you know, you've got chief people officers, heads of teleacquisition,

[00:24:31] [SPEAKER_03] and they're putting business cases together to go to the head of cybersecurity and head of IT compliance. They wouldn't always have to go there, you know, if they're doing a brand new ATS implementation, then probably. But they're just adding features and functionality, maybe not. So it's a new world for them. So we'd very much like to create packs for people and go, there you go. And if they've got any questions, send them to us and we can go through that in that detail.

[00:24:59] [SPEAKER_03] Because I think there's a lot to unpack. Does it work, number one? Is it delivering value and return on investment? Critical. We're all trying to do more with less. But is it doing it in a way that keeps you, you know, on the right side of ethics, on the right side of regulations, whichever that may be that you've got to adhere to?

[00:25:21] [SPEAKER_03] So, yeah, for us, that's the way everybody should be building AI in a very compliant way to give that support and guidance from day one.

[00:25:32] [SPEAKER_01] I guess I think of Olio as more than an ATS, right? Like it's certainly more than a traditional ATS, right? There's just a lot of, it's not just, you know, smarter, you know, features and, you know, AI this, AI that. It's an integrated suite, right? So are there some, I guess, I guess I'm curious how that plays into how you compete in the market.

[00:26:01] [SPEAKER_01] Some would argue, you know, if you're competing for dollars, you know, for budget, then that other vendor is a competitor, period, regardless of, you know, feature parity. But how do you think about just in terms of competitive intelligence and the capabilities that you offer? It just, to me, to the untrained eye, it seems difficult to sort of put people into boxes these days

[00:26:29] [SPEAKER_01] in terms of where you sit in the ecosystem. So how do you think about that from a competition landscape?

[00:26:35] [SPEAKER_03] Yeah, I think that's a great question. And it's a really interesting one. And kind of, you know, when you think about talent acquisition market, I think you have to make sure you're bringing in ERPs, you're in the big HCM platforms because they are able to do a form of recruitment offering. Is it as rich as specialist ATS platforms? No, but a lot of organizations want simplified technology stacks. So the market has become even more crowded.

[00:27:03] [SPEAKER_03] Then you bring in AI and you've got specialist startups that, you know, are focused on a particular area, particular challenge that you're trying to overcome. So it is a heady space at the moment. And we've done two things quite deliberately at Olio when we've been looking at how do we innovate with AI. So the first one is how do we improve on our infrastructure and what we have today and add real value?

[00:27:29] [SPEAKER_03] So how do we bring AI in a meaningful way that really helps to solve those challenges that people are looking at? So whether that's, you know, identifying the right skills for the role right at the beginning and start trying to find a unicorn and really refining and making sure they, you know, are expressing the skills and the experience they need in the right way to attract good quality candidates to, you know, further down the process simply being how do I take an interview note?

[00:27:59] [SPEAKER_03] How do I get better interview notes and provide really strong feedback to our candidates? So looking holistically at that process and saying, where can we add value on what already is a strong infrastructure? But then when I, when we look more broadly at the market, you know, trying to go and get enterprise organizations who, you know, to take out an ATS and put a new ATS in that's, you know,

[00:28:23] [SPEAKER_03] tens, you know, the cost and the upheaval and the time is huge for these organizations. So we've looked at, okay, how can we go in and help them leveraging our expertise, leveraging our compliant solutions, but in a standalone way. So how can we integrate into big systems such as Oracle and really add value to the customer by overcoming a specific challenge? So one area in particular is around screening.

[00:28:52] [SPEAKER_03] Now, we're all getting more and more applications. We're not quite sure if the candidates are being genuine, but we don't just want keywords. We want to understand the true skills of those candidates. So the way we've developed that screening tool is that it can ingest from a ERP, do what it needs to do and take it back. So for a recruiter, their workflow doesn't really change.

[00:29:16] [SPEAKER_03] The look and feel doesn't really change, but the speed and the accuracy that they can screener is hugely more efficient, times reduced, and they're able to understand those candidates. So we've really kind of tried to look at and go, if we integrate with these guys, what's the right way to do it? And how do we actually support the recruiters to get more value from their overall technology stack?

[00:29:41] [SPEAKER_03] And that for me, I think is a key area for us to be able to compete in kind of, you know, the top right corners are always the magic corner for everybody to be at with the big analysts, but to allow us to be in those conversations with these enterprise organizations and say, hey, look, we can work with what you already have today. You don't need to rip everything out. We can complement your technology stack, but help you solve that problem that you've got, that your current technology can't do.

[00:30:10] [SPEAKER_03] Long-term view, that will, you know, open up doors. We'll be delivering value. But for now, that's how we see ourselves competing in the market, really leveraging our ability to be flexible with other technology.

[00:30:24] [SPEAKER_01] No, I definitely agree. That's important. And I was surprised. Just saw a sneak peek at some upcoming research from Aptitude around ATSs. And there actually is a decent amount of respondents to their survey that are actually looking at new ATSs. So you got a lot of opportunity there. And people, I guess, are not afraid to, you know, put something or take out a system like that

[00:30:53] [SPEAKER_01] and put something new in if they see that incremental value that you're alluding to, Haley. So I think that's important. I think nothing, there should be no sacred cows, right? You've got to really understand what's your medium and long-term sort of vision for how this all fits together.

[00:31:10] [SPEAKER_03] I think AI has kind of opened that up a little bit, you know, because people are quite open to express their challenges a bit more than we were when it came to business challenges and recruitment challenges. So, you know, we've certainly seen a shift when we see RFPs and things like that come through. So the way the challenges are expressed are called out far more explicitly. I need you to help with this because they're feeling it acutely because of candidate usage and things like that.

[00:31:38] [SPEAKER_03] So I think, you know, you kind of, you need to be watching the market very closely to understand because it can shift very quickly at the moment as well. You know, what was a big challenge yesterday could easily by next week be completely blown out of the water by something else or some new technology that's entered the market. So it is moving at pace as well. And like I said, there aren't any sacred cows and there are lots of new players

[00:32:03] [SPEAKER_03] that are kind of challenging that status quo, which I think is good for organizations. And it's certainly keeping, you know, longstanding organizations on their toes. You know, where do you want to be competing and kind of where is your focus?

[00:32:18] [SPEAKER_01] One of the things that that brings up is the broader organization. When we think about, you know, AI adoption, you know, obviously across talent acquisition, across HR and well beyond into the organization across domains and lines of business. And so there's different sort of cohorts across those populations that you need to, you know, account for.

[00:32:46] [SPEAKER_01] And so I just wonder how you guys think about that or how you may work with, you know, other parts of Olio in terms of how some of the philosophy and the ethos around what you're building sort of translates and sort of jibes with what some other leaders might be, you know, attempting to build.

[00:33:11] [SPEAKER_01] And are they all held to the same sort of standards in terms of like, you know, responsibility and ethics and all that?

[00:33:19] [SPEAKER_03] Yeah, absolutely. And it's really interesting actually because I'll talk a little bit kind of about what we're doing in Olio because AI literacy, you know, certainly within my sphere is something we've been working on for a long time as we've been starting to develop and making sure that we're all, you know, learning, know the regulations, know how we need to develop, what are our new processes, what do we need to do on top of what we're already doing, but critically how, you know, appropriate usage of AI.

[00:33:48] [SPEAKER_03] But actually in the wider organization, that's a really big project for us, you know, because it's quite easy in my world to assume that everybody knows how to use AI. So we've got all kinds of, we've got like, we use the Google Suite as an organization. So we've got all kinds of learning pathways at the moment where people can go in and join in all of these Google sessions. But we've then got one-to-one, you know, led by kind of subject matter experts within the organization

[00:34:18] [SPEAKER_03] going to all departments and doing those sessions and those practical sessions. So a little bit like I spoke about, you know, doing it kind of for your own personal use, but kind of relating that technology to a task and helping them to understand, you know, how do you use a Gemini gem? How do you use a notebook LLM? You know, all of those, how would you use Claude? Whatever the tool is that you're using. But in a, one, a safe way, you know, people really understanding what we mean by no IP.

[00:34:48] [SPEAKER_03] They're like, oh yeah, there isn't any. And we go, is there a username? Is there a password? Like, yeah, we're like, that's like, you know, that's IP. So getting people to understand those sides, but making sure, you know, where, you know, what are their security credentials and what model are they using and what, you know, getting people to understand their basic checks they need to do. But then how can they safely use that in a work environment? And people love it. Like we have like lunch hackathons. They think it's great.

[00:35:18] [SPEAKER_03] And I've learned lots from my peers where they're like, oh God, I use it for this. And I think, why have I not thought about using it for that? I've got a little tool at the moment, anybody that uses Gmail, I've got a flow. So every morning in my calendar, it gives me all of my hints and tips that I need from my, from my emails. You know, what are my actions? Who do I need to respond to? What haven't I done? What have I been chased on? What's just for information? So it's all streamlined.

[00:35:45] [SPEAKER_03] So I'm not there kind of 9am every morning frantically trying to go through my emails and catch up where I've been doing wonderful things like podcasts the day before. But we kind of take that ethos into our client base as well and say, okay, let's run some masterclasses. We run a thing called the AI Forum. And that's all about education. I'm not going there to talk about my latest products or why you all need to buy it from me. I'm there to kind of, you know, Bob, you joined me.

[00:36:14] [SPEAKER_03] Martin Redstone's joined me. Lots of industry experts, but equally we bring our clients. They're already utilizing AI so they can kind of give their hints and tips on how did they manage to break down those cross-organizational barriers? How did they win the hearts and minds of people within their organization that, you know, maybe were like, absolutely no, no AI. These guys are going to say, look, I really think it's a good idea.

[00:36:41] [SPEAKER_03] So how did they break down those barriers and start their adoption journey? And kind of how far have they got? So it's a bit of an education as well, you know, and I think that's really key. You know, don't be scared to ask people and attend these sessions because you learn so much just by sitting and having, listening to those conversations of how somebody overcame it. There'll be a nugget where you think, I've experienced that. That would probably help me.

[00:37:08] [SPEAKER_03] And they're kind of safe spaces and safe conversations so you can ask all those questions that, you know, maybe you're not completely comfortable with kind of one-on-one. But you can kind of, you know, hide behind the chat and ask a question or slightly controversial question just to spark a debate. So you can kind of really learn kind of the edge cases of morality and ethical usage of AI because there's always a bit of a gray area with new technology.

[00:37:35] [SPEAKER_03] So that knowledge sharing, I think, is really key. But I would say, like I said right at the beginning, like our role as a vendor has certainly changed. You know, we've always worked at earlier on three pillars and one of those is trusted advisor. But that's more true now than it's ever had been. You know, our organizations are looking to us to say, how can we do this? How can we adopt AI in a safe way? And how do we break down those barriers?

[00:38:05] [SPEAKER_03] So, you know, we're learning as well. Every organization needs it in a slightly different way at the moment. But we're learning kind of what it takes to overcome those barriers easily.

[00:38:16] [SPEAKER_01] Well, there's a bunch of threads to pull on there. Yeah, go for it. The first, I'll try to keep them straight in my head. The first is just the communal aspect of learning, right? We all have knowledge to share. We all go off on our tangents and down our rabbit holes. And so to be able to bring that back and have the, you know, people kind of abuse this term,

[00:38:39] [SPEAKER_01] but psychological safety to just share, you know, failings, but not in a negative way, failing as a learning opportunity, right? Here's what I'll never do again. I'll never, you know, either use this tool or ask this question in a sort of naive way or trusted to, you know, have this particular piece of information or, you know, whatever it is.

[00:39:06] [SPEAKER_01] Someone else can be your sort of canary in the coal mine and take those learnings back. But, you know, share, being able to share successes as well as learnings with a community. We're all on a journey at some point in this, you know, sort of AI upskilling journey. And we all need to be ready. I mean, I use the term AI readiness quite a bit.

[00:39:30] [SPEAKER_01] And I just think that whatever your lot in life, whatever, either, you know, blue collar, white collar, you know, it doesn't really matter. You're the work that you do will in some way be impacted by AI. And so you just have to figure out, you know, what is the right balance point for you,

[00:39:54] [SPEAKER_01] but also recognize that, you know, different organizations have different sort of risk tolerances. And what you don't want to do is just ban it. I know, you know, I think Oleo works with regulated, you know, industries and things like that. And so there's a lot of sensitivity, certainly, but there's a lot of benefit. But this is the point of the responsible AI, you know, literacy is so that you know, you know, good from bad,

[00:40:24] [SPEAKER_01] you know, from acceptable to not acceptable. You know, where's the ethical line? And where's the line for your company and the mission of your team? And in your case, you know, how does this get interpreted by candidates? How does this enhance the candidate experience? How are we showing, you know, respect for them as an individual?

[00:40:48] [SPEAKER_01] And so there's a lot that you can get from, you know, those education sessions, but also making sure that it's, you know, open and transparent so that those learnings can go, you know, far and wide and people can name some that.

[00:41:05] [SPEAKER_03] Yeah, a hundred percent. And I think, you know, when I think in particular to one of our organizations, Police Scotland, they had exactly that scenario. You cannot use AI. And if candidates use AI, they've cheated. Yeah. And that, and this is kind of just over 12 months ago. And my goodness, were they determined in that talent acquisition team? And, you know, they helped shape the OleoQ product. So that it was, you know, really going to overcome the challenge.

[00:41:33] [SPEAKER_03] I mean, the volume of emails they were getting over a weekend, people, lots of questions. You can imagine you want to join a police organization. The number of questions and complexities in the process and the things expected of candidates is much deeper than some other sectors. But, you know, they were able to overcome through pure determination, I would say, initially to get the product in. But in a matter of weeks, the impact was there.

[00:42:01] [SPEAKER_03] So they were able to show that return on investment, you know, a 30% reduction, I think, from memory in their email inquiries, just in kind of three to four weeks. So they were able to be like, look, look at this. Look at how much time we're saving. Look at how we're now able to spend more time supporting candidates, understand what's coming in their medicals or getting them through the complex vetting process.

[00:42:26] [SPEAKER_03] That was a real kind of time drain for them within their processes. But that's transformed that organization. They're now, you know, now on a real strong AI adoption, not just in talent acquisition, across their organization. Because through one tiny little chatbot, people were able to, one, understand the technology and how it works far more than they were previously.

[00:42:51] [SPEAKER_03] But they grew a confidence because the impact was tangible and they could see it and they could feel it. So that is always my advice. Start with something that has an impact. It's not overly complicated. There isn't high risk on the AI compliance scales. And just think about, that's a problem we have. Let's try and demonstrate the value of this tool very quickly and then use that as a basis in the foundations to expand your journey with AI.

[00:43:20] [SPEAKER_03] Because those simple tools can be the, you know, can make such a huge difference. But they can be the light bulb moments in organizations that are really risk averse. Because a chatbot feels quite friendly, right? And you don't get in the perpetual loops anymore, Bob, because they're AI. You don't just get there. Speak to agent. Can't help you anymore. You know, but it can make such a difference by having that impact so quickly.

[00:43:46] [SPEAKER_01] Yeah. I think for modern AI that's using, you know, generative AI, I feel like chatbot is antiquated for that. At least that's my association personally, like chatbot. If someone says chatbot, I feel like they're talking about, you know, the earlier generation of assistants, you know, in the corner of a website. For example, that, you know, following very rigid, you know, commands.

[00:44:14] [SPEAKER_01] I can help you if you ask one of these five questions. And that's the extent of how I can help. Otherwise, I'm transferring you to a human agent. But the other, just quickly, Haley, the other thread I was going to pull on was just around like if you ban it or you say you can use AI,

[00:44:33] [SPEAKER_01] but only if it's this, this one tool, you're sort of asking for, excuse me, you're sort of asking for a shadow AI kind of problem, right? So for those not familiar with that term, we had shadow IT where people were using their own devices, their own software in a corporate environment. And now we have basically the same thing popping up again with AI tools.

[00:45:01] [SPEAKER_01] And there you, so if you ban them, someone's going to go on their personal device on, you know, potentially on your network with your data and get the job done that they know they can get done, you know, faster, better. And so you've just got to be careful. Again, it goes back to this theme around responsible AI.

[00:45:21] [SPEAKER_01] You just sort of, your inability to sort of trust people to use it responsibly, perhaps because you haven't given them the education in the beginning, your inability to do that doesn't increase the trust that people have in technology or in leadership's decisions about technology. And I don't know, I've seen this before, right?

[00:45:50] [SPEAKER_01] We saw this with social media when I was at IBM. People were like, oh, well, organizations are going to ban it because people are wasting time on social media. But, you know, that was before they realized there's a lot of B2B applicability to what social media afforded. So I guess my point is, you know, you can't avoid it, right? And your people aren't going to avoid it.

[00:46:14] [SPEAKER_01] So let's work together and figure out what the right policies are and make sure people respect those policies and we can all trust each other.

[00:46:24] [SPEAKER_03] Yeah, absolutely. And I think it's, you know, you hit on a really interesting point there for me about, you know, if you say you can only use one tool and you can't use another, that shadow AI can still be a part of it there. And I think that's actually, you know, an interesting point. People become strangely fond of whichever tool they're using the most. So on their smartphone, whatever tool they're using the most.

[00:46:47] [SPEAKER_03] And actually probably don't understand that, you know, a lot of the tools that are available in corporate suites, you can utilize in a very similar way. It's just they've not used it. It's not as familiar. So that peer-to-peer, that education, that communal approach to kind of AI literacy becomes even more important for organizations. You know, when I think about it here, like I said, we use the Google suite kind of in our day-to-day business world.

[00:47:15] [SPEAKER_03] When we're doing our developing worlds, they've got access to other tools, but our day-to-day with Google suite. And we've definitely had those conversations where people are like, but I just want to use OpenAI. And we're like, well, we're a Google house. We have all the Gemini accounts. It's within our, you know, it's our account. What are you looking to do? And they say what they're looking to do. And we're kind of like, okay, let me sit down. Let me show you how you can achieve that with what we've got.

[00:47:41] [SPEAKER_03] But it's that, again, there's like a fear of having to learn the new tool and they don't really know how to approach it. And they've probably stumbled across how they're doing it on OpenAI versus Gemini or whatever it might be. So, yeah, I think it's really important organizations take that time and kind of say, okay, if we want people to engage and embrace this and not do shadow AI, what do we need to do to give them the literacy skills within our tools to achieve what they're looking for?

[00:48:10] [SPEAKER_03] Because, I mean, it is great. It's great for productivity and efficiency. I love it for brainstorming personally. I think it's, you know, those blank page moments. It's amazing what you can kind of come up with and start to kind of to and fro. And you might end up with loads of different chats all over the place, but you can then bring them together. So from a brainstorming perspective, people haven't done it for that either. Then try it. But, yeah, education, I think, is critical. And doing it in a safe way.

[00:48:38] [SPEAKER_01] No, I like your idea of giving people the appropriate path for the use case or for the objective, right? And we see every day it seems I see a Substack post or a Medium post or something on LinkedIn around, oh, well, you know, OpenAI is doing this that we don't like. So if you're in there, here's how you switch everything from, you know, ChatGPT or Codex to Claude

[00:49:07] [SPEAKER_01] or here's how you do this in Gemini or whatever it is. And I think it's people are always going to move. They do it all the time. They switch, you know, phone providers. They switch cable providers, whatever it is. That's never going to stop. But I think if you show people that, you know, all those markdown files, all those skills that you built in one environment,

[00:49:30] [SPEAKER_01] they're actually pretty easy, easier than you might think to port over to a new environment. So let's move you out of this, you know, sort of risky, you know, sort of personal use situation and let's recreate that without a ton of friction into this new environment. Yeah, there might be some couple days or whatever it is to sort of get re-acclimated,

[00:49:58] [SPEAKER_01] make sure all your connection points are in place. But there's a better and safer way to do it. And I think sometimes I just need that personalized in the moment, you know, in context kind of education.

[00:50:10] [SPEAKER_03] And people don't necessarily always understand the risk. You know, how risky can it be? I always remember it's one of the first horror stories I heard about AI and that. I think it put people off for months, if not quarters, you know, and that was certainly, you know, I think it was Samsung and somebody went on and wanted something and uploaded a load of files. And that was all their IP because they put it on the public domain, you know. And it was just, they were just trying to do something.

[00:50:40] [SPEAKER_03] They were trying to be kind of forward thinking, how can I do this, but didn't understand what that upload meant for them. So, yeah, really, it's always really interesting to try and think about how do we do that.

[00:50:55] [SPEAKER_01] So, Hayley, without revealing anything about your playbook, what are you thinking about for, like, the, you know, the evolving role of the recruiter as this technology advances with more AI, more agents and agentic workflows and things like that? What are just some of your thoughts around, you know, what you're hearing from your teams and your customers

[00:51:22] [SPEAKER_01] and how you sort of, you know, scope that?

[00:51:26] [SPEAKER_03] Yeah, absolutely. And I think one of the key things to say is, you know, human interaction is going to be key. Humans in recruitment, you're irreplaceable. I think it's, you know, there is a level of human touch, humanity, and being able to read between the lines that AI, you know, isn't going to be able to do.

[00:51:52] [SPEAKER_03] It's not going to be able to pick up on human behavior in the same sense. So, you know, I think hopefully it's come across. I love AI. I think it's extraordinary and it's a great advancement in many ways across all organizations, but I feel particularly passionate about it in talent acquisition, you know, patterns, extracting that evidence, really helping you to kind of understand skills gaps as well as what candidates have within their applications. But for me, it's, you know, cultural alignment.

[00:52:23] [SPEAKER_03] That's becoming more and more important when we look at the world of work. You know, particularly with younger generations coming in, their need to understand culture from day one and know how they align. And equally, how as organizations do you ensure you're bringing in the right people for your culture? So, you don't have a revolving door because it's easier than ever to kind of move from role to role, I think is going to be a real kind of real focus.

[00:52:53] [SPEAKER_03] So, you know, making sure we talk about the kinds of experience, don't we? Making sure we use AI appropriately there. Because candidates want in the moment kind of real-time information, but there are aspects where a human needs to be involved. So, making sure we get that balance right. There's a real juggle at the moment, I would say, and how do we get that balance right between delivering in the moment information,

[00:53:19] [SPEAKER_03] hyper-personalization, which is a word that's starting to get overused, but actually having that real human interaction. So, I think that's going to be a really interesting space for us to be looking at and thinking about. But I also think that when we think about the world of AI, it's going to be interoperability. That's going to become really key. And I see kind of, for me, kind of 18 months, maybe slightly earlier, the way we think about our technology stacks is maybe going to change.

[00:53:48] [SPEAKER_03] So, how do we want to interact with our ATS to our Slack channel to our Teams meetings to, I don't know, your Power BI and your insights tools? And I feel those worlds will start to kind of merge. And that world of HR, talent management, talent acquisition, I feel them coming even closer than they have been before. And that strategic voice getting stronger in organizations.

[00:54:19] [SPEAKER_03] My advice to recruiters is be AI literate. Think about how you leverage AI within your role. And really start to upskill yourself so that you can, you know, manage knowledge bases. If you're looking at using chatbots and, you know, how can you be over the top and under the stand, the compliance risks and evaluate and do candidate appeals.

[00:54:44] [SPEAKER_03] So, think about how you can evolve your own skills because AI is here to stay. And the more you can do and the more literate you are, I think the stronger your voice personally will be when you're at the strategy table, when the worlds combine a little bit more than they already are.

[00:55:00] [SPEAKER_01] Yeah. No, I think that's great advice. I mean, you'll also be able to have a better sense of whether you've got a real candidate on your hands or not. And real answers.

[00:55:09] [SPEAKER_03] Exactly. There are still some telltale signs.

[00:55:12] [SPEAKER_01] We'll have you here for another hour if we go down that rabbit hole. So, we're going to park that for another time. Hey, thank you so much for spending some quality time with me. It's great to see you and always lovely to talk to you.

[00:55:25] [SPEAKER_03] Absolutely loved it, Bob. Hopefully, I'll be back again soon and we'll carry on. We'll go down that rabbit hole next time.

[00:55:31] [SPEAKER_01] All right. That sounds good. In the show notes, I'll put a link to your LinkedIn profile to Olio's site. And if you have a link to the AI forum series that you're hosting, I can include that too. Perfect. Perfect.

[00:55:44] [SPEAKER_03] Thanks so much, Bob. All right.

[00:55:45] [SPEAKER_01] Thank you, Hayley. And thanks everyone for listening. We will see you next time.