How Google DeepMind uses AI in recruitment

AI is speeding up recruitment – but can it really replace the human touch? Not quite.

In this episode of TA Disruptors, we go behind the scenes at Google DeepMind (one of the world's leading AI companies) with Becky Pradal-Rogers, Head of Talent Acquisition.

In a world where top talent is rare, roles are evolving, and AI is everywhere, Becky shares how her team expertly uses AI day-to-day – and where they deliberately refuse to hand over the reins.

From using AI to map talent pools to running interviews where candidates are encouraged to use AI, Becky reveals how Google DeepMind balances efficiency, innovation, and the kind of human judgment that algorithms can’t replicate.

This episode will help you cut through the AI hype and see what “good” really looks like inside one of the UK’s leading AI labs.

Join Becky and host Robert Newry as they unpack…

🤖 AI as support, not substitute – How Google DeepMind uses AI to map skills, understand talent markets and speed up admin, while keeping recruiters firmly in charge of shortlists and hiring decisions

🃏 Finding wildcards in an AI-shaped market – Why AI is great at spotting obvious skills, but still misses “wild card” candidates who don’t fit the mould yet could be game-changers – and how human judgment and institutional knowledge uncover that hidden talent

🧪 Designing for candidate AI use, not fighting it – How Becky’s team flips the usual “Are you using AI?” interview question on its head, by asking candidates to use AI during exercises and assessing how they think, prompt, verify and explain their work

⚖️ Responsible AI and real-world governance – Why AI literacy and due diligence are now non-negotiable, and how organisations can stay on the right side of fast-evolving regulations (including the EU AI Act) while still innovating

📊 Turning data into decisions – How AI is transforming recruitment reporting by consolidating data sources, surfacing trends and generating usable insights that help hiring managers move faster without cutting corners

🚀 Experimentation as a strategic skill – How Becky’s team carves out time to learn, test and iterate with AI, turning curiosity into measurable impact – and why that matters for finding talent for jobs that don’t even exist yet

❤️ Why recruiters aren’t going anywhere – In the end, AI can accelerate the process, but it’s still human connection, context and curiosity that win the hire

If you’re trying to work out how to embrace AI in recruitment without ending up with a team of clones, this episode is for you.

Listen now 👇

Transcript:

Robert: Welcome to the TA Disruptors podcast. I'm Robert Newry Co-Founder and Chief Explorer at Arctic Shores,  the task-based psychometric assessment company that helps organisations see more in people. 

And in this fourth series, I am focusing on the disruption in recruitment brought about by AI  and how TA thought leaders are responding to this technological shift. And today I'm very excited to be welcoming a TA leader from an organisation at the forefront of artificial intelligence advances, Becky Pradal-Rogers from Google DeepMind. And many of you listening will be familiar with Google DeepMind, which is the world's leading artificial intelligence research laboratory,  focused and founded  here in London in 2010, and it's done many great things over the yearsvbut there are two that stand out for me. And as a board game enthusiast, the first was the development of AlphaGo, which in 2016 beat the legendary Lee Sedol in a five-match game of Go, which is widely considered so complex as to a computer program beating a human would provide the next frontier in AI development.

And as somebody who grew up and witnessed and watched IBM's Big Blue beat Garry Kasparov in 1997,  I have been following the developments in this space. And so it was very exciting when Google DeepMind crossed that frontier and created a program at AlphaGo that not only beat Lisa Dole, but actually went on to beat any other type of system and program that had been developed in that space. And then the second area is also a personal one. My daughter is a geneticist working at the Francis Crick Institute,  also here in London, and she talks excitedly about AlphaFold,  which made significant advances in the predictions of protein structure. And it's the breadth of what Google DeepMind does that I think is so impressive.

And we are in lucky and we were talking a little bit about this before the start of the podcast to have that capability and that investment here in London here in in the UK and it's something that we don't always appreciate and celebrate that with so much  of the world focused on artificial intelligence that we have this incredible capability  right on our doorstep as well. 

As head of talent acquisition, Becky, you have this incredible job of designing strategy and execution for hiring across AMIR and APAC into this mind-bogglingly, uh if that is a word, advanced organisation, which must be both exciting and daunting. And we'll explore a bit more of that through the podcast. But very excited to have you here. Welcome, Becky.

Becky: Thank you very much, and thank you for having me. I'm very excited. It's actually my first podcast I've ever done. I'm excited to kind of get stuck in.

Robert: Well, I'm excited to have you as a newbie as well. And I'm sure this will be a very interesting and thought provoking session. Let's start with the point that Google DeepMind is at the forefront of artificial intelligence.

And so you must have in your DNA in the organization people coming to you and talking about how do we use artificial intelligence in recruitment. And it's probably different for you as it is for many of the listeners here where there is a sort of constant and quite loud messaging coming from corporate head offices where we must use AI in recruitment seems the perfect place to be using it.

And so it'd be really interesting if you just start off and sharing with us how we're using AI in recruitment. And I'm assuming it is in your DNA and you don't have to think about should we be using AI? It's just, it would be an automatic area for you to explore.

Becky: It's such an interesting conversation. I guess just a couple of things before I jump straight into the question. So one, you're going to hear me use the acronym GDM. So Google DeepMind.

Robert: Google DeepMind.

Becky: And I think the first thing is the reality of AI is that it can be really overwhelming. Again, just before we started this, we briefly touched on the conversation of, know, will it put my job at risk? So I think there's a concern that people's head go there quite early on. But then the kind of  pieces, there are so many options and AI has been a space that has accelerated so quickly and actually what seems like a quite a short space of time, the reality is organisations like GDM have been working on this for years and years and years, but the market has suddenly kind of shifted its direction and we're seeing so many AI tools in the recruiting space that perhaps didn't exist before. As a recruiter, it's...

Robert: It can be quite daunting as well, as there's so much...

Becky: Overwhelming of... 

Robert: Of technology, what it means. Where am I using Algorithms suddenly, and you're having to become a  statistician as well as a technologist, as well as a recruiter and all these...

Becky: Exactly. And so,  yes, it is in our DNA, but similar to other organisations, it's really about understanding what is the intention and the purpose of how we're using AI in our recruiting practices.

And so for us, the focus really is on like how are we using AI to be as efficient and effective as we possibly can. AI is not there to replace a recruiter, it's there to support and supplement in order for recruiters, and whether that's a recruiter, a sourcer, and in a broader ops role within the TA ecosystem.

Ensuring that their time can be focused on the most high impact work. So whether that's spending more time with your candidates, more time with your stakeholders in the organisation, whether that's hiring managers, HR partners, compensation teams, where are we using AI to give the team that time back? And so it's not a kind of case of saying, we're gonna use these three tools and they're gonna do a very specific thing because that's gonna reduce our cost in this specific area is actually thinking about on a day-to-day basis, where can we give our team time back? 

And so for us, we are really lucky in the organisation that we are in, that we have access to an amazing suite of tools across the kind of Gemini, Notebook LM ecosystem. And so for us, it has been a bit of kind of trial and error and experimenting of where do we want to use it? And as I said, it's about kind of efficiency and effectiveness. So I can jump into it. Some examples of what that

Robert: exactly. Let's have some of the examples because I think that's what people are struggling with. As you said earlier on on this one, it seems to be so broad as to how you could use it. And I really like your point about, well, you know, people talk about human in the loop and keeping it effective and efficient. These are big, broad terms, but actually we've got to start with something and where and how do you experiment and what have you experimented with?

Becky: Yeah, so I think three areas I kind of cover in my role are, yes, the recruiting side of things, but also the broader operational element. So that covers onboarding and actually our kind of ops is how do we scale TA systems, things like that.

And so if I start with, I guess, the recruiting side, we've seen AI accelerate the speed at which candidates can apply for roles.

Robert: Yes, and you've really seen that as it's a, you're getting inundated.

Becky: I think all organiations are probably seeing the impact of that. And so yes, you can look at tools where they screen CVs for you. That comes with risks. And actually for us, it's about better understanding where's the best source of candidate coming from. And actually if we can focus the team's efforts  on intentionally mapping the market, identifying this kind of top talent. And you can use, whether you use a Gemini or there's lots of other tools out there, whether that's kind of ChatGPT or... 

Robert: So you're using those tools for market mapping, as it were… research.

Becky: Exactly, yes.

Robert: So rather than a recruiter having to go and work out what kind of skills do we have in this area, where might they be, could we even access them,  which would take, I imagine, time and you may not even have access to the research, you can go and send an LLM out there as it were with a prompt to go and do that.

Becky: Exactly and a great example this week actually, a member of the team was presenting, sorry, last week is, had a fairly brief meeting with a hiring manager to understand the needs of the role, then actually was able to go away with that kind top level summary and research the market.

Find me the best research scientist, citing academic papers, publications, they want to put together a job description, again, putting in a kind of standard template, putting in the information that they kind of come out of that briefing meeting. And actually in a fairly short space of time and with bit of, know, kind of prompting here and there, shifting things, work that would have taken hours and hours can be consolidated into a couple of hours. You can then go and meet with your hiring manager. These are all the candidates that I've been able to identify based on the information versus scrolling through Google Scholar or LinkedIn pages. You can use the AI to help you go deep into that space.

So that's an example in recruiting where it's instead of spending hours trawling through CV screening, it's how do you pivot that time to being able to do effective market mapping, researching, passive talent in the market. And particularly in this space now, this is hot talent that a lot of people are going after. You can't rely on them to come to you. You've got to invest that time going after them. 

Robert: And just on that point. And I think that's because there is a subtlety in what you're saying there too. Because you're talking about research and market mapping. I assume what you're not saying is uh Gemini, go out there and find candidates for me. Because then you're in the domain of  how do you know what candidates  Gemini has selected. What reference points it's made. We know that the training models behind these LLMs is based on the great unwashed of the internet. And so we don't know what bias sits in all of that.

Becky: No, exactly. And that's where the human element comes in of recruiting, which I think is such a key part of what we do is  that human element. It's being able to look beyond the obvious. And I think one of my favorite things when I was actually a recruiter was that wild card candidate where you'd see that profile and be like, this doesn't fit  the mold or the norm, but I know there's something about them and I'm gonna advocate for them.

So that still comes into this. And yet I'm not saying, Gemma and I would find your list of candidates, here you go, hiring manager. There is validation, you're kind of understanding why has it identified this candidate? What are the papers that it has,  that individual has been part of um creating and the publications, but what it is able to do is to really consolidate that search phase into a much shorter space of time. And then you as the recruiter are able to kind of dig into it, validate and identify are these suitable candidates for my roles?

Robert: And just on that note, because I love the wild card story and it's something  that I've used in the past  to highlight the challenges of being overly CV-centric in our research and sourcing, is where does the wild card sit in this? So how do you ensure when we're using AI for all this research that there's still the opportunity for the wild card candidate to come up? Because the danger is if we rely on AI to do all the narrowing down on this, the wild card will never match the word prompt search that you put in?

Becky: And I think it's such an interesting question. And the reality is I don't have a perfect answer to that because I, you know, as a human, we have our own biases. I can't quite remember where I read this, it's always kind of stuck with me as a recruiter will spend on average five seconds, reviewing a CV before kind of flicking onto the next one, so we have our own biases of where we're looking for keywords or companies or how long have you been in this organisation for?

And so something I think AI will be able to do is look at skills, job descriptions, key criteria, and hopefully make some kind of recommendations rather than decisions  that can help us in this space. Do I think we're there yet? I don't know. And I know we were briefly speaking about this earlier. I think there is a lot of risk, particularly in this space. But I think one of the things as a recruiter is sometimes you might look at a profile, whether it's you find it on LinkedIn or you see it screening a CV and there's just something about it that will stand out, that will make you say, I'm gonna chat to that candidate.

Or maybe someone in the business comes to you and says, I've got this person, this referral, please just talk to them. And so you have that conversation and that sparks something. And it goes back to my agency days. I remember if just there is something about this candidate I've got to advocate for them. I know they don't quite fit, kind of, don't tick all the boxes. And so I think that human element of a recruiter is so important and I don't think you can...

Robert: And we don't want to automate that out. And that's my point around all of this too,  and that's the efficiency and the effectiveness. If we go too much on the efficiency scale, we lose the effectiveness, which is sometimes the wild card candidate that is the one that actually is perfect for the reasons as a recruiter that your instinct, your knowledge built up over many years of the organisation of people. 

Actually, there's great value in that knowledge and capability and just human perception. It's not necessarily bias, actually, that is just good human perception.

Becky: Yeah. And you build up, I'm so lucky I have an amazing team that I work with and you really realise that that institutional knowledge that they build up is what really helps them to identify like what is a good candidate for this specific team and it's understanding the makeup of that team, gaps in skill sets, where have people come from before? What can we bring to the table that there's a value add here, not just a kind of carbon copy of what we've seen before. And I do think that is one of the risks that comes with relying solely on AI if you're looking to... uh give time back to the team, you've got to identify at what cost. So I think it's a really interesting question that I don't think there was a perfect answer And we're still figuring that out.

Robert: And I think it's interesting that even at Google DeepMind, you're trying to figure this out like the rest of us, because it's pioneering. We are at the forefront of change here  and we won't necessarily know all the answers, but all that we can do is to make sure that we're asking the right questions in the right way. 

And on that note, you made the point about risk in all of this and understanding risk.  And I think that's a key element in all of this. So what are the questions you're asking of your team when you're thinking about do we use AI or how do we use AI? There must be there clearly are some risks around this, whether it's systemic bias, compliance, there's another risk in there. So how are you managing some of those risks?

Becky: Yeah, again, it's a super interesting question. And I think we're all navigating this space. It's kind of ever evolving. We've seen things with the EU AI Act coming into play. As an organisation, we take you know, responsibility in AI incredibly seriously. So actually something we all do across the team is training in how to use AI responsibly.

Robert: Oh, so you have specific training on that.

Becky: specific training on it to ensure.

Robert: Developed your in-house around.

Becky: Yeah, so we understand, you know, of um implications of actions that we  might take. I think it's really important that we also self-educate ourselve a talk which I think you came to a couple of months back where I didn't actually realise it's not the responsibility of the provider that you are using. So if you're using a third party provider, it's not their responsibility to make sure that the technology meets regulations.

Robert: That's right, it's you as the employer recruiter as it were.

Becky: The organisation purchasing that product which I thought was a fascinating piece of information that I hadn't been aware of again, if I think of my role specifically, if our ATS system is rolling out a new AI feature, I take the time to understand with the provider, what is the feature? How is it working? If you're using an enterprise solution, which I imagine most people are.

In most instances, it is not training on your data. is not storing it, you know, if it is, it's a very short periods of time. But I will always work with our legal team to understand. Does this stand? Are there any risks  before we will make the decision to kind of go ahead and utilize that feature?

Robert: So you'll bring in the legal team. mean, it's standard anyway, that we have to turn to do a data protection impact assessment anyway. But that I'm, that's interesting that you'll do that with your ATS provider, who I know is greenhouse and an organisation that I have a lot of respect for, particularly what they've done about responsible AI too.

And that's an interesting point because there's quite often people do a data protection impact assessment when they're introducing new technology, a new vendor, but not necessarily when a feature arrives from a vendor and the whole work day versus maybe a thing in the US has, I think probably changed our view as you were saying that you are responsible for making sure  that this feature that's arrived  is not actually going to create a new risk  for you. so that every time there's a new feature you will be bringing in and checking.

Becky: I will always check with the legal team. And the reality is, know, as an organisation, in most instances, you would be working with a legal team when you were implementing new systems, signing contracts. And so you'd want to make sure that  you understand the terms of that system that you are using and how your data is going to be used and how your data is going to be protected. But yes, personally,  even if it's just a check-in, I will always take those steps with the legal team to kind of make sure that any steps that we are taking, any features that we're turning on.

Robert: Compliant and not generating any risk. Because there's two elements to this risk. One is there is just a compliance one, so the legal team do that. Then the other is  a, and maybe part of the compliance is around fairness in there, but sometimes you, yeah, how do you work out that you are making decision making in a way that is  fair before you implement it? you, is that just a question of having to ask the… technology vendor that they've done the quality checks around that because  you don't want to be suddenly finding yourself saying, oh, it appeared to be compliant, but a year later now we're seeing the data coming in and we've had a lot more men  that have been selected rather than women.

Becky: Again, I think there's this piece around being curious. So I going back to your original question, how are you using AI and so forth, I think a really important part is getting curious and understanding how these tools work, how data is trained, how are you then going to apply that to your organisation or your process or whatever it might be. And so having that understanding and asking those questions, I don't think you should just rely on, oh, we'll kind of punt that over to the legal team because I think.

Robert: Because sometimes they'll understand the legal, but they won't necessarily understand the technical element, then those things are different.

Becky: And so I think it's just important that you yourself understand how something is kind of operating and working and then you can make decisions  that suit  your organisation. But yeah, I would say we are in a very lucky position that we have the infrastructure around us to ensure that we are being thoughtful and mindful about how we're taking those steps. I appreciate for smaller organisations, they may not have that. And so again, it's… being curious and understanding how to kind of equip yourself with the right resources to be able to make those decisions. So important. 

Robert: And so we talked a bit about how you do that for sourcing, but you also said earlier that you are  being impacted in the way that many organisations are by candidates now having the same artificial intelligence technology and capability in their hands through a computer or a phone - probably more so than organisations now. And so that's  in, when you're, and I assume you have roles where you open it up to the market, it's not just sourcing. And so you're facing this sort of deluge of applications.

So how are you thinking about what technology or AI can do to address that are you going down a kind of deter and detect route around this or yeah, how are you  addressing that challenge that many people, this kind of sea of saneness, tsunami of applications problem?

Becky: Yes, so again, we are in exploratory phase here uh of really kind of understanding impacts, implications, efficiency, effectiveness. So it is an area that we are exploring more. Again, it's about asking the kind of questions of how is this data being trained? And I think you yourself, you have to understand, will your kind of previous hiring trends, when utilising one of these tools, will it give you outcomes that represent that or will it give you outcomes that

Robert: That actually just embed a world where there was already bias previously.

Becky: Yeah, exactly. So this is like full transparency, something that we ourselves are exploring. I think another  area which is very much related to this is, you you hear this kind of question around how a candidate's using AI in the interview process. Are they kind of, uh you know, cheating the process? And again, I think that there's some really interesting tools out there that, you know, talk to this.

When we were already thinking through this question, was like, what if we flip the question on its head and it wasn't about how do we identify if candidates are using AI to support them in their interview process?

Robert: Really interesting.

Becky: And actually it's like, how do we use AI in our interview process with them in partnership? And so what we've been piloting recently in our coding interviews  is  an interview where we asked them to use  an AI system of their choice.

Robert: You expect them to be using it...

Becky: It could be Gemini, it could be another organisation's product, but we will ask them to use it in a coding setting. And so it's not then worrying about  are they putting these questions into 

Robert: some sort of tools surreptitiously.

Becky: instead it's, well, let's have you use it. Let's have you use it in this interview so we can see how you are utilising it. And the reality is, AI is evolving in a way that it is gonna impact all of our roles, but that can be done in a really positive way. So how are you, Coder One, using AI to enhance  your ability in this role? And it's actually been, it's fairly early days in the pilot, but it's been really interesting and it's actually helped us to identify or kind of validate some things that came out of… other parts of  one of the other coding interviews.

So again, it's not just about how do you use AI to screen CVs or check if somebody's cheating, but how do you think about using AI in your processes in a collaborative way,  where you are working with the candidate versus using it as a kind of way to-

Robert: Oh, to catch one out. Fundamentally take the starting point, which is using AI which I always find totally bizarre, that on the one hand we're saying, come into, doesn't matter whether it's Google or many of the other organisations out there that are saying,  you're gonna be using AI as the way that you work, but you can't use it in the recruitment process because we think that's cheating. And it just seems mad. So I think, is that your kind of view of the  trend, the way recruiting is going is… you might as well just embrace AI because actually if you see it in a positive way, you can use it in a way  that helps you find better people rather than trying to exclude people.

Becky: And even in the interview process, if you're looking to bring a recruiter onto your team, it's how are you using AI in your role to make you more efficient, make you more effective? And so why would we not apply that to other roles that we are hiring for? how can we use this technology in a really positive way in our roles versus using it as a way to kind of say, you're cheating or let's screen you out of this interview process. So it's early days.

Robert: So it's early days, you're just pilot phasing at the moment. So at the moment, do you provide any guidance on your website about good use and bad use of AI for candidates? No, we don't at… at the moment, it's a very good question.

Becky: Something to note to self in 2026. We are actually doing some work on our careers page, so watch this space. But yeah, think it's such a kind of fast moving, evolving space that I'd love to know if anybody right now has nailed it and is able to give like the kind of… a perfect example of where they're utilising AI that's kind of demonstrating ROI in a way that...

Robert: fair for people and not putting them off  and all those things,

Becky: Yes. I think there's so many ways. And another one just really briefly, as I said, because I cover the ops space as well, is  helping people to self-serve. onboarding is an area as well that we cover and… there are so many questions when you're coming into a new organisation, particularly that the size that we are and sitting within the larger Google ecosystem.

So we actually use Notebook LM to build a self-serving agent of, I've got questions about my onboarding experience. That is your first point of call. Or in our TA side of things, again, so much information, processes. And so we've just done a big audit consolidating all of our information. And again, putting it into Notebook LM, the team can now self-service, they can get questions.

So there are all these little things, but if you actually kind of add up the time that you're saving, you might find that you've saved two hours in a day, you might say that you've saved five hours. I've been using it a lot in data analysis and I may have spent a day doing very manual data inputs. And now, using Gemini Insheets, Google's version of Excel, I've been able to teach myself some fairly basic, coding is not the right word. No, formulas.

Robert: Formulas.

Becky: Which would have taken me hours. And now I've kind of built the confidence to be able to do that through utilising this tool. It's not necessarily doing the work for me, but it's helping me understand, well, actually, I mean, it's building the formulas for me. And then I, from that can… you know, analyse the data, I can ask it questions to kind of help prompt me on certain things. So I think there's many ways you can use it.

Robert: Yes, there are many ways. And I think that's kind of interesting on the data analysis side, because, you know, a huge part, I'm sure, of the work that goes on at Google DeepMind is about data analytics and how you can use artificial intelligence to support that. And that probably feeds then into data being a big part of recruitment at Google. I think data is increasingly an important part of  any  organisation that's doing a large amount of, or is global and doing a large amount of recruiting.

So I really like that idea that you, how do you use AI to help you in  self-serve in some of that data analysis? Because we're flooded to some extent with data from our applicant tracking system, from other third party tools we might have around that. And so it's interesting that you're saying, well, actually maybe we can use AI to help us in consolidating. And is that how you're doing it to kind of generate reporting around this too and help you with some of your analysis?

Becky: Exactly, can be weekly or bi-weekly reporting to the organisation. Yes. So it can help just consolidate different, you know parts of data, pull it together. There's an infographics feature that we can use again, instead of you having to make all these fancy graphs and whatever it might be, it will pull that together for you could be,  you're looking at a load of different data sets and my brain doesn't naturally go to, I can see the themes that are coming out here.

It can very quickly help you to kind of identify some of those themes and whilst it's not giving you an answer, it's helping you to kind of question things and dig into parts of the data a bit more. for me, actually in my role as a manager, I found it incredibly helpful from the data analysis piece and actually building my confidence in, well actually I can do a bit more formulaic work in sheets or...

Robert: I like that confidence point too because I think that's true for a lot of us that you...you're familiar with Google Sheets or Google Docs, but you know that there's this enormous capability in those applications that you don't touch because you've never really had the confidence to, and you know it's gonna take you hours to, I mean, look up formulas and that kind of thing. I know once upon a time, I did know how to do it, but I kind of...know now that it would take me hours if I wanted to go and research it, whereas I could just get an AI tool now, Gemini possibly, to just say, right, create me a formula and is that how you're using it now?

Becky: I mean, this may be a very embarrassing thing to say, but I didn't even know how to do conditional formatting until that long ago.

Robert: Well, that's right  was probably something we may have come across briefly, but we know we've been properly trained on it.

Becky: And if I actually would search how to… how to do conditional formatting, I'd find it a bit technical and give up quite quickly. Whereas through the use of AI, it's break it down for me in simple terms, create me a formula. And so it goes that one step further. And then from there, I was like, I'm actually, this is now interesting. So I reached out to someone from my analytics team and was like, I'd love to just spend half an hour with you to help build my knowledge and my capabilities in this space.

And again, our analytics team are amazing and always so accommodating. So they sat down with me  and then what I was able to do is actually build some kind of formulas which now service me on a weekly basis. Whereas before I would kind of sit there and manually, I've got to pull from here and I've got to pull from here.  And  something that would have probably taken me a good day's worth of work. I can turn around in an hour.

Robert: Wow. The other thing that I love about that story, think it's that confidence that we need in TA,  which is  many people were probably thinking, gosh, there's Becky  in Google DeepMind. You've got some of the  greatest brains in the world sat out there with data analytics. You'd almost go, could I possibly reach out to them and ask them some help? Because I'm just going to be embarrassed. It's like asking a Formula One racing car driver, can you help me just start my Mini? And you just think they're going to think I'm mad on all this as opposed to actually do reach out.

They're probably really happy to help you learn and understand. And we've got to kind of get over that  barrier that we have. It's entirely personal barrier in there that I'm going to feel stupid. And so I can't possibly ask this question or make this request for help. 

Becky: Absolutely. And I think the reality is, I might lean very heavily on a team historically because there are certain capabilities I don't have and I haven't had the confidence to lean into.  Actually, I'm giving that team time back by no longer having to  pester them with, could you do this for me? Could you help? Whereas I've used the tools available to me to help build up that confidence and understanding  to the point I'm then like, oh, actually, hey, member of the analytics team, and also in this instance, wonderful Lydia, can you help me expand upon this?

So now I can self-service more. And as I said earlier, this point around our onboarding  capabilities, it's how do we help people to self-service, which then gives team members time back that they would have spent answering questions. So it's all these little things that when you add them together and you kind of look at the portfolio of where you've applied AI across your processes, it really can make a real difference to the outcome at the end of the week. And actually, just something else on this note, we did quite recently, probably about three weeks ago is we dedicated a whole morning to AI training. So there was a number of different sessions that were run. You could kind of opt in for which ones.

Robert: And this was just in your team, was it, or was this broader?

Becky: This was more broadly across Google, which we all took part in. But it was saying to the team, like, we want you to prioritise this. We're kind of saying, like, please do take a morning to lean into this space. Give yourself permission to, you know…

Robert: Experiment.  

Becky: Even after some of those sessions, I reached out to the trainers to be like, I'd love to understand a bit more about this. Can you help build my knowledge base? But I think it's giving yourself time. Like these tools, you're not going to instantly pick up an AI tool and be like, I know how I've mastered this. I can use it straight away.  You have to invest a bit of time understanding how to use it, understanding how to do good prompting.

Robert: So you've got to carve out the time for that because, and that must be hard because you know you're familiar with it will be for most TA teams on this that most a lot of the time you are firefighting almost in the sense that you just got a backlog of roles or something urgent has come in and so you're not always master of your destiny on this. So I suppose does it help that you've had a corporate directive that says we're gonna carve this out or are you also making time in your own diary for your team to do this?

Becky: Do you know what, it's both actually. So it's great that the organisation has said,  this is something we want all of us to kind of focus on, but it's really important that then you, I think as you know, as your kind of smaller sub teams invest in that as well. So since then we've had follow-ups,  we've had team members present back to the wider GDM team on this is how I am using AI in my work and this has been the impact and the outcomes of it. Can I help you in this space? 

So I think it's a kind of bit of both, you have to, when you first start out in a recruiting for a role you've maybe not worked on before, you're not gonna get through 100 CVs in the same amount of time you're going to. You've kind of trained your brain in knowing how to, know this space, I know what I'm looking for. And I think it's similar with AI, you've got to invest in understanding how to use it to benefit and see the impacts in your area.

Robert: Right, be curious, experiment, invest in the time to give yourself to think about it and experiment to some of the key things around this. get that. And where is your mind going for 2026 then? Because you've shared with us brilliantly some of the little areas where you've bought back time. And so what are the big challenges for 2026 and how are you thinking that technology might solve or support?

Becky: In all honesty, it's building upon the foundations that we've put in place this year. This market is, you know, the intensity is ever increasing

Robert: when you say this market is for the type of things like data scientists.

Becky: Yeah, so you know, the AI space. So the thing with this is it's such a new  industry in so many ways that roles that didn't exist two years ago, now exist. There will be roles in the coming years that don't exist now that will exist. And these talent pools are small. There may be people that you're looking for and there are… handful of them in the world, or there may be people who've never actually done this role.

Robert: That's right, so you've to go out broader.

Becky: So you have to be really thinking about where do skill sets translate. These people might not be sitting in an obvious company in London, they might be sitting in some niche market somewhere else across the world. And so for us, it's how, if I think about next year, how do I really get my team to be able to hone into finding, connecting and hiring this very specialist, highly sought after talent and to be able to do that takes time. And so it's how do we find that time?

And I'm somebody who's kind of a big believer in not asking the team to do more with less, but actually really looking at how are we operating and how can going back to this, know, kind of efficiency and effectiveness to give ourselves time back.

So again, really looking at our processes, how we're using tooling and systems to be able to really hone in on that high impact work and try and move as much of the kind of administrative burden that we can. But again, it's about which we were talking about earlier is there are going to be trade-offs and there are going to be things that you have to really consider of. If I want as a member of my team to be able to give more time to candidate engagement,  what needs to give to be able to do that? And actually really thinking about the trade-offs and the risks that might come with that.

Robert: Yes. And just on that point about trade-offs  and  the requirement  to find people outside of your traditional very narrow skillset. How are you thinking about that? Because I've been a big advocate for in situations like this of almost scrapping the CV because if you've got a very narrow pool in there and there aren't enough people with a specific skill set, you've got to find people outside of it.

So the CV is not going to be helpful around that. So, how are you thinking and exploring  that you will be able to  move beyond that narrow skill set and broaden out your search criteria?

Becky: I think there's a piece in this around networks, and really understanding from  your internal teams, like who do you know in this market? Who is somebody that might not seem like an obvious fit, but is someone that you know that could kind of transition into this role. We're incredibly lucky that we have insanely amazing people in the organisations who have these brilliant networks and have come from all over the world, different backgrounds, know, it's kind of academia, it could be more in the kind of like, you know, politics, government sector. And so it's really kind of tapping into that more. It's understood from hiring managers. What else could this person be doing right now?That translates into this.

Robert: It's challenging the hiring manager a bit, isn't there?  Because they probably know that they may have come from a different route, academia for something.  But it's very easy to get into that mold, as it were, saying, well, we've got to have somebody that has experience in this rather than, well, actually,  what is it about you that brought you into this in the first place?

Becky: And again, it's conversations with those hiring managers around the must haves versus the nice to haves. But again, these people aren't just going to land on your lap. You have to be able to give the time and investment to understanding the market where these people might be. And so it is this, as I mentioned earlier, it's not just posting a role and hoping that people apply to it and having to sift through endless CVs. 

It's actually really spending the time to engage with the business, with key stakeholders, with the market. It might be actually somebody that you hired a year ago, tapping into their network. I have a couple of… members of my team who would be actually better placed to answer this question, because that's something that they spend a lot of their time focusing on is this niche talent. And they do a phenomenal job at finding it through, know, broadening out their searches, their networks, tapping into the kind of ecosystem that we have.

Robert: Yes. And I think that's, you know, such an interesting point that we, because the talent pools are so narrow on this, that we have to think about and take a different approach to finding that talent. It starts with the hiring manager and the must haves and the nice to haves and the things that we used to think were must haves being put into the nice to haves and the things that we used to have as nice to have being put in the must have, which can be soft skills, for example.

And, and so I, you know, like that point is that you know, we've got to  think about these things a bit differently. And on that note, then you've articulated so well the value that recruiters are going to provide in this. It's going to be  the knowledge, it's going to be the research, it's going to be the understanding of both the marketplace  and the organisation.

So what's your take on some of these conversations that have been coming out from a few of the AI companies around, the AI is gonna take our jobs. Will there be a job for a recruiter in five years time?

Becky: I hope so. uh I genuinely believe that  such a key role that we play as recruiters is that human element, is that human skill. You will remember that time when you dealt with a bad recruiter and that time you dealt with an amazing recruiter. And so that part of always understanding what makes a candidate tick?

They might come to you at the beginning and say, it's the team I get to work with, it's the research I get to work on, it's that prestigious individual who has  worked  on  alpha fold. That is what excites or drives me, or whatever it might be. Fast forward to the latter parts of the process, is that still the case? Has something changed? And I don't think technology is able to replace that human element of really understanding what makes a candidate tick and what's going to make them choose your organisation over another.

I do hope technology is able to support us in making fairer decisions, moving at greater pace, removing heavy administrative burdens. And so do I think it'll impact our space? Yes, but I really think that we… we as recruiters have that special something, which is that human interaction and knowing what makes a candidate kind of tick that can't be replaced.

And even if I think about my agency days, I didn't even have a laptop. We sat at a desktop, we had paper CVs, someone would call out a job and you'd get up and you'd wave a CV in the air. We've already seen technology evolve in terms of,  you know, ATSs and, you know, tools that might be transcribed for you or...  So we've seen it impact our roles a positive way. I think that will just continue, but it's understanding how you, the part you see technology playing in the evolution of your TA function. And I guess you can embrace that and lean into it, or you can shy away from it. But I think we know the inevitable is that technology's not going away.

Robert: It's not going away. It's going to be more important for part of our life. So we need to embrace it.  So very, very good thoughts and advice.  One final, final question on this then. You talked a lot about, you know, starting your life,  you know, in the agency model,  CVs.  Do you think we have heard death knell of the CV now that AI has just made it so easy for somebody to enhance a CV or a cover letter that actually the information that we're getting from a CV is not something we can rely on anymore now and that we're going to have to look at other ways  to understand  somebody's expertise and capability.

Becky: I know you're, am I right to this, you're not a...

Robert: I'm not a fan of the

Becky: You're not a fan of the CV, are you? Do I think it's the death... Not yet. I don't think it is. But I would hope that there is more to come in this space that we're not pinning someone's future career on On a bit of paper. I would love to see this space evolve more. And I generally think that AI can play a really important role in that. 

But I think perhaps we revisit that question… in a year's time. 

Robert: We can come back again and discuss that. think that's a great point.  Becky, it's been fantastic talking to you. Lots of great thoughts and insights. And I know a lot of people would have taken some great value from that. So thank you for coming on to the podcast.

Becky: Thank you so much for having me.