AI Takes The Boring Part w/ Adam Godson

The Agentic Future of Talent Acquisition with AI

In this episode of HR’s most dangerous podcast, Chad Sowash and Joel Cheesman sit down with Adam Godson, the newly appointed General Manager of Talent Acquisition Products at Workday (following Workday’s 2025 acquisition of Paradox). The trio dives into the rapidly shifting architecture of recruitment software, the looming "SAS apocalypse," and how artificial intelligence is changing the actual nature of HR roles.

Key Discussion Topics

• Workday's acquisition of Paradox

• The Shift to an "Agentic" Architecture

• The "SAS Apocalypse" & Evolving Revenue Models

• Data Ownership and Candidate Passports

• The Reality of Recent Layoffs

• How AI Re-frames the Recruiter's Job

"We take a 3D person that's got a whole life, and we flatten them into this one or two page text document called a CV or resume... and then we take a job which is multi-dimensional and complex, and we flatten it into a poorly written job description... and we try to match the words on them thinking it's going to come up with something. That's no way to live."— Adam Godson

PART TWO - AI Takes The Boring Part

PODCAST TRANSCRIPTION

Joel Cheesman: All right, let's do this. We are the Chad and Cheese Podcast. I'm your co-host, Joel Cheesman. He's Chad Sowash, and we are here with Adam Godson, GM of Paradox. Adam, welcome to HR's most dangerous podcast.

Adam Godson: Joel, it is so good to see you. It's great to be dangerous here.

Joel Cheesman: And nothing says dangerous like Midwest milquetoast over here, Adam Godson.

Chad Sowash: Here we go.

Joel Cheesman: Here we go, let's go. A bunch of Midwesterners hanging out.

Chad Sowash: Yeah, no, no kidding. GM, CEO, Paradox. You get acquired, that's awesome. We'll talk about that. But this new role, but it's more than just Paradox.

Adam Godson: That is true. So I think, you know, of course Paradox was acquired by Workday in September '25. And, a few weeks in, they're like, "Hey, you seem pretty smart. Why don't you also run Workday recruiting and HiredScore, messaging, candidate engagement, and all of the talent acquisition products?" It's really like a focus on that stakeholder, which I'm excited about, which is having a unique stakeholder for talent acquisition that, for Workday clients, so we can really make solutions for and putting those things together to be sort of Workday recruiting and the sort of the agentic future of what this is going to look like.

Chad Sowash: That's a big portfolio.

Adam Godson: Yeah, it's having a great focus on talent acquisition.

Chad Sowash: Yeah, that's a big portfolio.

Joel Cheesman: Yeah. So, so who, who talent-wise from those organizations are part of your team now? Is this going to be a house of brands? Is it all going to be a Workday recruiting thing? Like, talk about sort of vision of brand and talent around that.

Adam Godson: Yeah, yeah, good, good question. So in terms of the sort of the story about the product, it is largely going to be about, you know, sort of Workday and then the agents to get work done. So, you know, you guys know like as long as there's been ATSs and, you know, TA technology, it's been my choice is I can choose the thing that's connected to the rest of my stuff, my platform system is connected to my job architecture and the data, or I can choose the thing with the features that I wanted, right? And so this best-of-breed model, but I have to integrate that in.

And I think the future now is different. I think you can have this connected model, but then have agents be able to do the work and run on top of those systems and get the features that you want and the connection to the system of the, you know, the rest of the data in your in your company. And so, you know, that for Workday clients, it's really about sort of Workday recruiting and its agents.

But one of the thesis behind the Paradox acquisition and the HiredScore acquisition as well was to continue to serve clients that aren't on Workday HCM. And so the Paradox brand will stay. So we'll have cool circles everywhere, so that tattoo I got is not going to like end up being awkward, which is...

Joel Cheesman: Uh-huh, yeah. We won't mention the tramp stamp on you.

Adam Godson: Yes, thanks for not mentioning that. But, but we'll continue to serve clients that are standalone Paradox, that were on SAP or Oracle or other systems, and continue to serve them really, really well as a as a way. And, you know, for me personally, that was really important as we thought about the acquisition to serve the whole market and to be able to not just serve Workday clients, but to be able to make a difference everywhere. Workday's the biggest, yes, but it's even still, you know, I want us it's a fragmented market. I want to be able to serve it all.

Joel Cheesman: Will, will the advent of agents make brand less important because who cares, the agent works with everything, and it's not about the brand so much?

Adam Godson: I think you're right. Like as the as the work surfaces change, and as people start to work in fewer places, it's more just GSD, like get stuff done. And so you don't care if it's this or that or this this brand or that brand. It's about what are you doing for me? And can I pay for those outcomes? And can I get the outcomes to happen in a way that makes sense for where I work?

Chad Sowash: Interesting. So are we moving from a model where, you know, we had these clunky applicant tracking systems, CRMs, HCMs, and really those are turning into systems of record, which we always called them in the first place?

Adam Godson: Yeah, we did.

Chad Sowash: But they're just turning into big filing cabinets, and then we're going to have an orchestration layer on top of them where you're actually feeding data into those files and pulling data out of those files. Is that kind of like the simplistic way of my...

Adam Godson: It's a little it is simplistic a little bit in that the...

Chad Sowash: I mean, who did it come from, of course.

Adam Godson: Those, you know, the systems do have really complex workflows. And like whenever I sit in a demo with like a complex global company, I'm reminded by like, "Oh wow, this stuff actually is really complicated," the number of flows that companies have. And so, but, but you're right. And I think of it like the backbone. Where like the ATS is going to be the backbone of the workflow and the job architecture and some of the things that are, you know, really core to the company. That's needed.

But then the agents are going to do the work. And they're going to do the the orchestration. They're going to, you know, um, they're going to understand when when work needs to be done, and call a sub-agent to go run a process, and then tell you when it's done, and file that in a log. And what will change is that humans won't be doing the clicking in those systems. Agents are going to be in those systems. Humans are going to be really valuable in engaging other humans and convincing them to join, and telling the story of the company and those kinds of things. But man, the work is going to change a lot. Yeah.

Chad Sowash: Well, it sounds like we'll be spinning up agents a lot for whenever we have needs, instead of spinning up features.

Adam Godson: Yeah, that's right.

Chad Sowash: We'll be spinning up agents. And those agents, prospectively, could be working in teams. So this is a this is a fundamental change. From an architecture standpoint, are a lot of these big systems ready for that? And I mean, obviously, Workday bought you guys for a reason, Asana and Workrise. Um, are we ready to to to get to that phase yet?

Adam Godson: I think the the the larger players are varying degrees of readiness for this. Um, and, and I would say I would put Workday in that bucket as well, in that there are many flows that are fully ready, and we're ready to go, and we can execute those flawlessly. And there are others that we've got to open up new APIs and have some some philosophical change in how the the data is um is stored and act accessed in order to to enable that that change.

Um, but I'll say that that, you know, Workday is on that case. And so that's a lot of the movement today is to say how can we be a more open system, how can we be uh ready for it to be an agentic foundation. But that foundation matters a lot in terms of the You can't imagine the complexity of like the security model and like the roles and permissions of these global companies where someone asks a question about some information they may or may not have access to, and to be able to to understand like what access that they have and what should I tell them, how does that blend with other information that they may or may not have access to. And those things get really complicated.

And so the the core systems are well-positioned to be a part of this, but they've got to frankly modernize to be ready for this agentic future. I think at Paradox we've got really good um early use cases and are driving ROI. And and as we bring companies along, it's like how do we find them the first two or three wins so that they're heroes in their own organization, and that their management saying like, "Awesome, go get me wins four, five, and six." And and that may take some development to to come, but it's really about like just getting started and getting the W's on the board.

Joel Cheesman: You've probably seen that the term "SAS apocalypse" is trending on social media. Curious about your take. How would you define it? Is it an overhyped concern? Do you think it's a real concern? Talk about your take on that issue.

Adam Godson: Yeah, I think I think um, you know, some of that has to do with like the market valuation of that. And, you know, market valuations are built on future expectations. And if you expect SAS margins to hold for the next decade, I think that's probably it's clear at this point that's not likely to happen. Uh, my 13-year-old vibe codes apps on the weekends now. Well, that didn't happen a couple weeks ago or a couple years ago, right? So like things are clearly different. So, um, I think um, uh, it's a reflection that things are are different. And I think it's it's really just a how are companies going to um adjust to that, and there's going to be new winners and new losers. And sorting out who is going to win and who is going to lose is the market's job.

And I think in in moments like this, it's it's people kind of get clumped together and the whole market kind of goes down, where I think the next phase of this is to say, "Okay, this is durable." And, um, and and then other things are going to be not durable. And look, the market would reflect those things individually in the next chapter.

Joel Cheesman: So assuming you what you're saying is it is a real threat?

Adam Godson: Yeah.

Joel Cheesman: So how is Workday and the the businesses in that in that universe preparing for that threat?

Adam Godson: Yeah, I think I think it's about how do we work in the ecosystem. And how do we um open the platform? I mentioned that, but uh uh also bring our own agents to market. And so that Workday is the the surface where people get work done. So we launched um um both our own native agents, but then also agent builders and a new surface there when someone has a question about how to do a workflow with in HR or finance, that they can uh use uh Workday to do that.

So that was a lot of what the Asana acquisition was about. It was about being that front door to work. And so if you have a question about um a projection of or a finance question, or something about how your um location strategy may be um impacts your finances, that that data is there and the access controls and and all the things that would need to answer that question are in Workday.

Chad Sowash: Yeah.

Adam Godson: Um, and so I think being able to have that has been a big deal.

Chad Sowash: I mean, at this point we take a look at SAS and we take a look at, you know, like the move forward to the anything agentic, and and the the big systems are already doing that. But the big differences is just literally the revenue models that they're putting in place. So it seems like the big shift is with revenue models, because the tech's already happening that way.

Adam Godson: Yeah.

Chad Sowash: So, I mean, is that really Again, the street's looking at this stuff, and they're they're kind of stupid because the tech's already moving that way in these SAS platforms, and the SAS platforms are the ones with all the data in the first damn place, right? So it seems like it's more of a revenue revenue model issue, which I mean, it can't be changed overnight, but but still, it's easily I think it's easily changeable.

Adam Godson: Yeah, I I think um, you know, man, I think we'll look back at the SAS era and say that was pretty good gig. Um, and and we're able to build one platform that was configurable and sell it to thousands of of of people. And um, I think there's a couple of vectors. One is is being able to to charge more for results rather than the subscription, right?

Chad Sowash: Or consumption or whatever you want to call it.

Adam Godson: Yeah, of course. And and then you, you know, just end up in a different different spot. And I think companies are going to have to adjust to how they think about um buying software, which, you know, some of them were overpaying, some of them were underpaying, and got to sort out who's who.

Joel Cheesman: So your so your thesis is essentially the model will change from a pricing standpoint versus big co America is just going to vibe code their own ATS?

Adam Godson: Um, I think I think the I mean, what you're saying is it is a real threat? That might be an extreme example. No, no, it's not an extreme example at all. Like like, um, the the the build your own software is going to be a um a pressure point, no question, about how how that's there. But when people talk about building your own, I often times say like, "Oh, you're going to build your own and then maintain it forever." Um, because that's the second responsibility.

Chad Sowash: We saw that happen in the mid-aughts.

Adam Godson: We did. We did.

Chad Sowash: And how'd that go for...

Adam Godson: I mean, I mean, very few companies in the in the in the last chapter were able to to build something of scale and then maintain it over any period of time. And then one went back to to other other products. I do think that the way that um software is going to be built in the future with AI is different and can provide new opportunities for both building and maintaining.

Um, I think some of the the the question is going to be about the the the core of those systems and when companies think about the very most important things for them, their people and their money, like who do they trust? And and is it going to be an in-house team that is vibe coded something or is it going to be the very, you know, the some of the the the core vendors that they're going to trust.

Joel Cheesman: Yeah.

Chad Sowash: Well, and I think that we've seen two examples, at least two, uh where CodeWall hacked Jack and Jill, right, with AI hackers. And then then Mithos is is pulled back because so if you want to be an amateur and vibe code your own applicant tracking system and hopefully put it up against, you know, an army of AI hackers, versus allowing the professionals to actually do their job. I mean, so so when you guys, and again, this has got to be a big part of the discussion watching this happen out in the news, uh this is I mean, obviously it's scary but it's also an opportunity.

Adam Godson: For sure. Well, I think that it it's about what is core, what is what is something that you do that is like what other companies do, and then what is net new? And finding ways for a company like us to enable the all those things. So, for example, if you, know, um, someone wants to use Workday Recruiting as their core because it's got the the security model built in and and and the global um uh roles and permissions that are going to matter for your your company, the separation of data so that people in Germany can't see the French and the French can't see the Germans and all the other things that you've got built in, again, the complexity is wild.

But then we've also got things like Workday Extend, which is the opening of our platform, which is to say great like if we don't have something for a weird and wonderful use case you have in like employee health or something like, cool, build your own app on Workday and access that data model, access uh those APIs to make your own app on top of that. That's totally cool. Um, and then, you know, uh be able to use um our conversational intelligence layer in Asana to be able to um ask the questions, "Hey, what did my pipeline look like for this?" be able to to get those updates in Slack or Teams and operate in natural work surfaces.

So, I think that's the area we're in of just figuring out the operating model and then getting the pieces of like what is the right tool for the right job in the different spaces, and and I think we think that Workday at the core belongs in the center of that. And then there's of course lots of things to build on that will help companies operate. Yeah.

Joel Cheesman: I'm sure you're familiar with the case against Eightfold, basically unveiling what is the algorithm, why, just similar to a credit check, "I didn't get the loan, exactly why," and creating transparency around that. I look at that and say this is not just a typical lawsuit that's going to be settled out of court. I think this has potentially real teeth. Wondering your thoughts, uh if you do think it's a threat, what do you think the end result is? Do we come to a more transparent world where these systems that rank candidates now have to tell you how they ranked you, the way that they ranked you, and why you didn't get the job?

Adam Godson: Yeah, I think broadly speaking, you know, transparency is is a good thing in this, and that helpful regulation can help clarify the next era for AI in HR specifically. And so, um, while some people would like to think, "Ah, regulation is inherently bad," um I think inherent like regulation can be really good for us to be able to understand what those look like. I think it would be helpful to have that not happen at the like state, city, and local level so that there are a million regulations rather than like, you know, some rational number.

But, um, look, the way this is going to, you know, play out is is uh some level of of inevitability of of people are going to use AI to um help reduce friction in labor markets, to be able to find help companies find talent, help talent find companies. And uh uh I think our job as practitioners who are helping in this industry is to understand how we can make this fair, transparent, um, you know, compliant with sort of human rights and sort of the you know the rights of people. It's it's not online shopping. It is something that's deeper. It is something that's personal and involves your livelihood and your self-worth and other other things. So I think I think our our job is to help make good in the world of that, but um, you know, it's going to it's going to be a journey to get there.

Joel Cheesman: So do we embrace that because I've over the years, I know you've heard it, people have actually talked about creating some of of like a credit rating and at least running parallels with uh with uh credit ratings. Do we just embrace that and say, "Hey, yeah, that's what we do, but we're going to make it transparent, and we're going to have the candidate validate as opposed to just go out there and scrape stuff off the internet and throw it into their their profile." Um, we're going to make it so there's a validation process and the person who owns the data, the candidate, actually does that. Should we embrace that?

Adam Godson: Yeah, I I think there's going to be uh a lot to embrace in the coming chapter. I think um we're on a time ticker of the time where human beings use web interfaces to apply for jobs. So some of the agent-to-agent things are are going to come where it's going to be...

Chad Sowash: Joel can't wait.

Adam Godson: How can my agent apply to your your agent? Um, but I think I think you've you've hit on some of the key points is how does a human in control of their own data, um how do we have transparency understanding of what decisions are being made about a human person, yeah, and and how do we um, you know, just understand how the system works and not have black boxes. Yeah.

Joel Cheesman: All right, I can't let the agent on agent comment go go un- un-questioned. Um, currently there's a real issue with uh a flood of candidates, some real, some not, um lazy apply, etc. Uh, on the other side of the the court, it feels like a game of whack-a-mole. How do we plug the holes? What's real, what isn't? Where does that lead us to? If it's agent-agent, is there a confirmation that the agent is legit and the everything's legit and then at what point does a human come into that equation? What does the future look like?

Adam Godson: Call it called blockchain. We miss we did miss one in that era. That is for sure. We we could have made some of that work. But, um, yeah, I don't know. Get all get all your tin hats, fellows, but, um, you know, I think I think this is going to centralize around identity, of of an identity around a person and a legitimate job.

And our goal as an industry is to reduce friction in labor markets to help have people those finding each other in a way that makes sense, right? And and the reality is the system we have today isn't very good. So we take like a 3D person that's got a whole life, and we flatten them into this one or two page text document called a CV or resume.

Chad Sowash: 1D or maybe.

Adam Godson: Right? And like that sucks. And and then we take a job which has like is multi-dimensional and deep and complex, we flatten into this poorly written job description that someone got from an old system 10 years ago and just hit like accept on, right? So we're taking these two lousy two-dimensional documents and trying to match the words on them and thinking it's going to come up with something. Right? Like that's that's no way to live.

Um, and and so so so I think the future is we've got to get more depth on both sides of that equation. We've got to get more depth in jobs. We've got to get more depth in people. And and there's going to be a lot that um uh uh agent can help us with in communication. One of the fundamental parts of that is got to be some limit to in the system of spam. And that's that's the downside is that we can create a world, and that's frankly what what I hope to do at our client board this week is to um work with our clients to create a better world.

If left to its own devices, we may just spam the heck out of each other, and employers spam candidates, candidates spam employers, and then everyone's mad. But I actually think that we can create a system where where maybe you can only apply to N jobs at a time. And you get pre-qualified to say like, "Oh, you're actually of a decent an okay fit for this one, you're in the running, and you're not in the running for these." And the agents do all that work and start to have some rational limit on both sides of the marketplace to say, "All right, the limit now is human attention."

So how can we give like some some rational set of of pre-qualification to the level of human attention and actually like make the labor market work so that people can get hired. And yes, maybe we're the agents are spamming the heck out of each other underneath the covers, but the humans don't don't know about it. And I think there's a world that we can create there that will work.

Joel Cheesman: Always the optimist.

Chad Sowash: Oh, well, he has to be. He has to be.

Joel Cheesman: Always the optimist.

Chad Sowash: Uh, so I remember back in the the the olden days uh with uh with uh Taleo when they had their universal resume. Remember? Uh, one resume for passport one one resume, yes. And then iCIMS talked about a passport way back in the day, and it's just like we have these great ideas that never come to fruition. They never get the the traction that we need. It feels like we're we're finally getting that, or we might be pushed into it by by regulation, whether it's here or in Europe, one way or the other.

Adam Godson: Yeah.

Chad Sowash: So, as you talk about depth, from a candidate standpoint, obviously transparency, you providing transparency to them, but also them owning the data that's on them, like the background check. Yeah, the company pays for the background check, but that was a background check on me, so that is mine, right? Are we going to get to the point where companies finally understand that they do not own that candidate information, the candidate owns their information, and at that point, can we move forward and finally creating whatever the hell we want to call it, universal, passport, are we finally going to get there?

Adam Godson: I think the what you said about push is right, where and we're starting to see some of that in certain parts of the world. So like Singapore has Singpass, Estonia in Estonia, UAE has sort of the centralized identity system. So I think as as um Worldcoin is trying to, you know, so scan your iris and confirm that you're a human. So so but but all those are efforts to solve some externality around distinguishing humans from agents and AI.

And and I think we can all see from here that this is going to become a real problem. And so I think this will all accelerate, and for our not just this reason, it is it is our job then to take some of those movements and contextualize them for the purpose of looking for a job and to say great can we um, you know, capitalize on the work that's being done so that people can open bank accounts and buy things and have an identity, but also like, "Oh, hey, us too, how about we apply that to employment and make so that useful there?" Because like we're the ones to do that. People will not naturally do it, we'll create terrible systems all over again and live with it for their generation. Yeah.

Chad Sowash: Yeah. So is is Workday a part of that process and prospectively being a part of a of a bigger cohort of building that kind of like ecosystem? I mean, or or these conversations happening now?

Adam Godson: Those those conversations are happening. I'm I'm not here to announce anything today. But I'll say that um that those conversations are happening, um just about how we think about um uh identity and how we think about how to make Of course, the application easier to to do that. But that's that's sort of a short-term um problem. But it is about how do we think about making the economies of our societies work better by reducing friction between people and livelihoods. Mm-hmm.

Joel Cheesman: If you were to create a keyword cloud of our podcast in the last year, one of the biggest words would be layoffs.

Adam Godson: Oh.

Joel Cheesman: I'll admit that I'm uh sort of undecided on where this goes. Um, AI is taking all the jobs. We overhired in COVID, and now we're just correcting. Um, the Fed interest rates are why people are laying off. I can literally listen to a handful of reasons and think, "Okay, that makes sense." But I want to know from your standpoint, explain the layoffs to me that are going to go on. What is to blame?

Adam Godson: I think it's all those things in some fashion, and um I do think AI will have a meaningful impact on changing employment. Um, and just like any technology that's been invented, we fought full wars over the loom being invented, if you're well, no one else remembers, but... But technology has changed like changes, and it makes massive impact in society, and and so I think um some of it is that is is uh the the re-skilling that needs to happen of people that um are impacted by AI versus the needs that companies have for inventing new things with AI creates um challenges.

Joel Cheesman: So if I could put words in your mouth, uh and correct me if uh you don't want me to put anything in your mouth. Um... Short-term, yes, uh Fed rates have a big impact, uh we did overhire for COVID and those are that's course correcting. But when we look back 5 years from now, we will say AI has had the greatest impact on jobs. And maybe not taking away jobs, but creating new jobs and re-framing what maybe recruiting was 5 years ago, recruiting is a different skill set today. Is that kind of what uh I'm hearing you say or no?

Adam Godson: Yes. And I think, you know, and if you take the recruiting example, for I do think recruiting is changing and it is, you know, going away from um I think some of the delivery things, clicking buttons in systems, and going towards people that are doing system design and development of systems.

I was at a in a room full of, you know, TA leaders a couple weeks ago and said, "How many people have a recruiting operations function?" Every hand in the room goes up. And 3 years ago, that was not the case. It was it was it was not that. And and sort of we get this split between um, you know, recruiting operations and people that are designing the the functions or using technology, they're using AI to to automate to make them more efficient.

But then also, you've got your recruiters that are talking to candidates, convincing them to join. Like AI can't do that. AI is not going to get you to move across the country to to relocate your family and those things. Like only a person can can do that. What you get rid of is the people that wish they were talking to candidates but are clicking buttons in the system. And that's a good thing. Yeah.

Joel Cheesman: Yeah. What didn't we ask you that you wish we had?

Adam Godson: I don't know, it's probably time for beer.

Joel Cheesman: And with that, this has been the Chad and Cheese Podcast. I'm Joel, he's Chad, he's Adam Godson, GM at Workday. We out.

Chad Sowash: We out.