Rebecca Warren and Michael Watson discussed the evolution of talent acquisition (TA) processes, emphasizing the need to adapt to modern AI-driven technologies. They highlighted the limitations of traditional TA systems, which often lag behind technological advancements. Michael shared his experience with AI interviewers, noting their efficiency in automating frontline screenings and improving work-life balance for recruiters. Rebecca stressed the importance of redesigning TA processes to integrate AI effectively, rather than merely layering new tools on top of outdated systems. They also addressed concerns about AI replacing jobs, advocating for a shift towards more strategic, human-centric roles.

Introduction and Speaker Introductions

• HR.com Moderator welcomes attendees to the webcast on the new hiring engine, agentic AI in modern talent acquisition, sponsored by Eightfold.
• HR.com Moderator explains the process for receiving HRCI and SHRM credits and encourages attendees to ask questions via the Q&A feature.
• Rebecca Warren introduces herself and Michael Watson, mentioning their roles and experiences at Eightfold.
• Rebecca Warren outlines her career journey from talent acquisition to leadership and her current focus on talent-centered transformation.
• Michael Watson shares his background in talent acquisition, his transition to Eightfold, and his role in building the customer success practice.

Challenges in Modern Talent Acquisition

• Rebecca Warren discusses the idea that talent acquisition (TA) is behind due to the rapid pace of change in the world, not because it is broken.
• Michael Watson agrees, noting that TA is often the first team to be laid off and the last to be hired, leading to a constant state of hurry-up-and-wait.
• Rebecca Warren reflects on the evolution of TA processes, from the introduction of ATS to behavioral-based interviewing guides, and how these processes have not kept pace with technological advancements.
• Michael Watson emphasizes the need to fundamentally change processes to keep up with the pace of innovation, using the example of the shift from traditional job boards to AI-driven recruiting.

The Role of AI in Modernizing TA Processes

• Rebecca Warren and Michael Watson discuss the disconnect between current TA processes and the rapid pace of innovation, particularly the shift from AI to agentic AI.
• Michael Watson explains the concept of AI friction, where AI amplifies existing process inefficiencies, and the need to redesign processes to accommodate AI.
• Rebecca Warren highlights the importance of not just layering AI on top of existing processes but fundamentally redesigning the flow of work.
• Michael Watson shares his experience with a customer who implemented an AI interviewer to automate frontline screenings, freeing up recruiters to focus on more strategic tasks.

Audience Interaction and Poll Results

• Rebecca Warren and Michael Watson engage the audience with a poll asking what best describes their current hiring process.
• The poll results indicate that many organizations are evolving their processes with a layer of AI but are not actively redesigning them.
• Rebecca Warren and Michael Watson discuss the amplifier effect of AI, where AI exposes existing process inefficiencies and amplifies them.
• The conversation shifts to defining different types of AI, including automation, generative AI, and agentic AI, and their respective roles in TA processes.

The Impact of AI on TA Processes

• Michael Watson shares examples of how automation has improved tasks like scheduling and employee referral processes.
• Rebecca Warren and Michael Watson discuss the importance of automating mundane tasks to free up recruiters for more strategic work.
• The conversation touches on the fear of AI taking jobs and the need for recruiters to adapt and embrace new technologies.
• Michael Watson emphasizes the importance of having frank conversations with customers about the need to redesign processes to accommodate AI.

Redesigning TA Processes for the Future

• Rebecca Warren and Michael Watson discuss the need to shift from a SaaS model to a service-with-software model, where AI and other technologies are integrated into the overall service.
• Michael Watson highlights the importance of choosing vendors that offer agentic AI solutions and are committed to continuous innovation.
• Rebecca Warren and Michael Watson discuss the concept of a human-in-the-loop approach, where AI handles routine tasks, and humans focus on strategic decision-making.
• The conversation concludes with a discussion on the importance of staying curious, humble, and open to new technologies to adapt to the changing landscape of TA.

HR.com Moderator 0:08

Welcome, everyone. Thank you for joining us for the webcast, “The New Hiring Engine: Agentic AI in Modern Talent Acquisition,” sponsored by Eightfold. This webcast has been pre-approved for HRCI and SHRM credits. Please be sure to attend the complete webcast in order to receive your credits. You will receive an email from HR.com within two business days; it will include the certification credit information. You may also log into your account and view the credits that you have received. If you have any questions during the webcast, please click on Q&A in your webinar controls and type them in. And now it’s my pleasure to turn it over to Rebecca Warren to get us started.

Rebecca Warren 0:53

Hey, hi, and hello, folks. Welcome to the Rebecca and Mike Show. Well, no, not really, but I am Rebecca Warren, and I have Michael Watson here. We are going to talk through the new hiring engine: agentic AI in modern TA. We’ve got all kinds of thoughts and opinions and questions. We would love for you to throw any comments or questions in the chat for us. We may get to them live, or we may follow up with them later on, either in an email or potentially in a LinkedIn post. So, Rebecca Warren—I have been at Eightfold over five years, and started in TA in my career, moved into leadership, worked for a lot of different organizations—small, large, consulting, full-time. Then came on over to Eightfold to help build out the customer success practice. And so I did that for a couple of years, along with my buddy Mike, and he’ll tell you a little bit more about that. And then my last two years, I have been in Talent Centered Transformation under the marketing team. So I spend a lot of time thinking about thought leadership and disrupting the talent lifecycle: how do we think differently, using AI, using skills, changing how we work for what’s happening in today’s world? So Mike, over to you to introduce yourself, and then we will jump in.

Michael Watson 2:20

Rebecca, thank you. Before I introduce myself, let me just share a quick story. It was about 20 years ago that I was running the marketing department for the San Joaquin Human Resource Association. I was studying for my PHR. I never got it. Things happen, and I chickened out. I stopped going to classes, but it really does feel full circle to, 20 years later, be presenting on an HR.com webcast. So my name is Michael Watson. I’ve been with Eightfold a little over six years now. And prior to that, I was in the TA space. So I started with Aerotek right out of college, spent about 10 years in the contingent space before I moved over to the corporate world, where I’ve run every single type of talent function over there. I started in sourcing, and I’ve done coordination, recruiting. I’ve run global teams, and that’s what I was doing prior to joining Eightfold. I was running a global TA team for a cybersecurity company, and this was 2018 when I found out about Eightfold. So I’m somewhat of a dinosaur in the industry now. I’ve been using HR and AI together for… yeah, “expert,” maybe, is the way you call it. What you all [call] the expert. Look, it’s all gray and it’s falling out. But happy to be here today. So I’ve spent the last almost six and a half years now building out the customer success team. Hired Rebecca. She was my very first hire, and now the last year and a half, I’ve been focused on customer agentic enablement, so getting our customers up to speed on using agents to help them be more efficient in their role. So happy to be here.

Rebecca Warren 4:06

Alright. So to get us started, Mike, you and I have talked about this many times, about how the world is changing so quickly and TA isn’t, in some cases, keeping up. So my statement here: TA is behind. And so when I say that, I don’t mean that, you know, things are broken, things are terrible. It just means that hiring worked for the work that we used to need to do. So the challenge is now speed; the world has moved, and processes have not kept pace. Workflows have not kept pace. So Mike, let’s talk a little bit about that. Like, you know, a lot of people say, like I said, “TA is broken.” I don’t think TA is broken, but are we asking the current process, workflow, team, etc., to operate at a speed that it was never designed for?

Michael Watson 5:06

So let me go back to the first part. You said, like, “TA is broken.” It was never broken for me, and I don’t think it’s broken now. But what happens is we’re the first team to always get laid off, and then we’re the first teams that they need to hurry up and hire when they have their budgets approved. So we’re always in this constant state of hurry up… hurry up and wait, if you will. Yep, yep. So by the time they say, “Oh yeah, we need to ramp up recruiting again. We need to hire all these people,” we are somewhat behind the eight ball because finance and sales and the organization, they all have their goals, right? And now they’re waiting on us in TA to get caught up. So I think that’s where… you know, and it didn’t hit me until right now, Rebecca, when you laid it out, that that was the problem that’s been going on. Now, what happens? I have to hire all these recruiters. I have to get them up to speed. I have to train them on what’s important to us, our core products, our differentiators. It just takes time. And that continuation of building these things on just adds to this anxiety, if you will, or this process problem.

Rebecca Warren 6:13

Yeah, so let’s dig into that. Yes, it feels like, you know, I was in TA before TA was a thing. I was in TA when we were called “Personnel,” and then we were called “Recruiting,” and then we went to “Talent Acquisition.” And now, I don’t know, “Experience,” “People Experience,” I’m not sure what the next role is. But I’ve been doing this for a long time, and so we created processes. Like, when we first got an ATS, we’re like, “This is amazing, and we don’t have to deal with, you know, paper resumes anymore. We’re going to scan them all into the thing, and they’re going into the box, and we’re so excited to have an ATS.” And so we created processes around that. And then we said, “Hey, we need a better way to connect to our candidates. We’re going to build behavioral-based interviewing guides.” Huge fan at the time, right? And then, “We’re going to connect it to our competencies.” Did that at General Mills; like, it was groundbreaking, and it was so insightful as a recruiter to start connecting candidates to values, right? So we created these processes and we created these workflows, right? Like, we’ve got in our ATS, you’ve got to go ahead and enter in your req information. It’s got to get approved, or maybe it’s got to get approved ahead of time, goes through all of these levels of approval, then you can recruit for it. Like, the process that we created was based on the systems that we had. So if I think about what’s happening now, we’ve got more tech than we’ve ever had, but I don’t think TA is keeping up with it. Like, it feels to me like people are saying, “Our processes are comfortable. The way I’ve done things is comfortable. I don’t want to change it. I was hired as a TA leader for what I know. Now things are changing. I don’t know it. It feels scary, so I don’t want to do that.” So there’s this resistance, and I think our processes were built for what we had. And in my opinion, we’ve got to blow it up a little bit. So I’ll just throw that back at you, Mike. We haven’t really talked about this—well, I mean, we kind of have, but not really. So what do you think about that in terms of the processes maybe slowing us down just as much as how recruiters are entered and exited in the system?

Michael Watson 8:24

We did. We talked about this a little bit, right? We talked about the pace of innovation, right? And it’s outpacing our processes. And what I mean by that is, go back 50 years ago: the whole process of applying for a job and interviewing didn’t change for hundreds of years, maybe thousands of years, right? When I started recruiting in ’97, ’98, Craigslist… the digital age was just starting. Yep. So I remember going to some of those recruiters that were a little bit older than me at the time, that were used to posting in the Sunday newspaper, right? And saying, “Hey, we have this new thing, this new digital Craigslist. We can post on there. We can change our process. We can recruit faster.” “No, no, no, Mike, I don’t want to do that,” right? So then we went through this digital… to your point, we went through this digital collection of resumes for 20 years, right? What happened? The first thing you did the last 20 years when you got a job requisition, you ran to LinkedIn. It didn’t matter what you had, who you recruited in the past, who you had in your ATS system, right? So once again, even though innovation is speeding up, you know, I’m still trying to fit it into these old-school systems, which don’t work, right? And so now, what’s going on, right? So it took us 20 years in the digital age. The last five years has been what we call the AI age, and now what’s going on? It took us five years to go from AI to now agentic. So what took hundreds of years went to 20 years, went to five years. And now who knows how fast this is going to innovate? But we cannot continue to use the same processes we use, right? For putting the “For Sale” or the “Help Wanted” sign in the window, expecting someone to walk in off the street, the little bell dingles, have a conversation with them, and I hire… Yeah, those days are over.

Rebecca Warren 10:19

Yeah, agree. Okay, so we clearly have seen that there is a disconnect coming in. Now, we’ll talk about AI in a second, about what that does to the processes that haven’t changed since we started using ATSs, LinkedIn, all of the different job boards, things like that. But I want to ask you, Mike: so my role is, I do work with customers. I work with prospects. I work with, you know, the industry… kind of work a little bit more externally. You’re working and training and leading customer enablement for all of the folks that are using Eightfold. So talk to me a little bit about where you might see the disconnect in where customers are at, what they’re doing, how we’re trying to get them to use an agentic AI platform. Like, what’s the lag that you’re seeing between where folks are and maybe where they want to be, or where they need to be?

Michael Watson 11:16

Yeah, listen. You don’t know what you don’t know, right? So that’s why a lot of times you rely on folks like you and I, Rebecca, to help them, right? So when we kick off something with a customer, many times that customer is coming in and bringing those traditional workflows to the conversation habit, right? The first conversation we’re having with them is saying, “Hey, if we are going to fundamentally transform the way in which we attract talent, hire talent, engage with talent, then on the flip side of that coin, we need to fundamentally change the process which supports it,” right? So, Mr. or Mrs. Customer, it is hard, but this is why we have those frank conversations early on in those design stages. So we don’t… you know, the worst thing we can do is try to retrofit a modern technology with your legacy processes.

Rebecca Warren 12:11

I remember those conversations. That’s what caused this to build, right?

Michael Watson 12:14

Don’t do it. Trust me, do not do it. Take the time to design new processes around that, right? A perfect example, Rebecca, if I’m going back into TA today, I’m leveraging the AI Interviewer for all of my frontline screenings. My team never again has to go and do a frontline initial screening. Ever, right? Because that’s where this is going. And you can either say, “Okay, I’m on board,” like I was, you know, eight years ago now, in 2018 when everyone said I was crazy. “Mike, you’re crazy. AI, you can’t use it in HR. You can’t do this.” Well, I guess I am crazy, because here I am, here we are. And just like I told you that seven years ago when I first started here, that that new trend was we’re going to use AI, we’re going to rediscover talent, we’re going to move faster… the new trend I’m seeing now is this AI Interviewer, leveraging that to take that off your recruiters’ plates.

Rebecca Warren 13:23

So let’s talk about that, because… and I was in the same boat when I started at Eightfold over five years ago, and I said I was working for Eightfold.ai, and they’re like, “Oh, you mean Eightfold.com?” Like, “.ai” wasn’t a thing; like, it was a whole different space, right? But here’s where I think we need to have… we need to have that conversation. It’s that “AI friction.” Because when we look at AI capability—and it is doubling every four months or even faster, right? That’s the latest number—you can’t just layer AI on bad processes and expect it to be better. Because AI is going to amplify whatever it enters. So yes, I agree with you. I am super excited about the agents that are coming in, what agentic looks like. However, I think we need to have a bit of a pause and do exactly what you’re talking about. Like, “Hey, we need to fundamentally change the process.” Because if you just throw… like, you can do an AI Interviewer on the front end, and it’s going to make a difference, but if you think about your whole entire process, your tech stack, you have to think intentionally about what that looks like. Because, as we said, that’s that AI friction: in theory, it should help. But if you just put it on things that don’t work great, you’re going to get something that doesn’t work great, faster.

Michael Watson 14:46

For sure, right? It’s going to expose the warts and the scars, everything that you’ve been trying to hide in your processes. Yeah. To pull the mask off. So that’s exactly right. Now, listen, the other side of this AI friction is: how do we… and maybe we’re going to talk about it in the slides… is, how do I talk about my teams? Or how do I talk about this with teams, so they’re not sitting here freaking out saying, “Oh my gosh, they’re trying to automate everything. My days are numbered. I might as well look for a new career. I might as well get out of recruiting.” Because that’s not what we’re saying, right? So there’s got to be conversations also with your internal teams about the AI and friction, right? What’s in it? Why are we doing this? You know, one of the things I told all my recruiters when we first implemented Eightfold was, “Team, if we do this right, I think this is actually going to give you work-life balance. You’re not going to have to…”

Rebecca Warren 15:46

Sort of that elusive thing we chase.

Michael Watson 15:48

Yeah, that BS term we throw around a lot. Yeah, it was real. I was like, “Man, if I can leverage AI to do some of this meaningless—or not meaningless, but just that reduction of mundane tasks, right? Allow my recruiters to be more of that talent agent.” I’m all for it.

Rebecca Warren 16:07

Okay, so hold that thought. That is… you are on the right track. We are going there. What I want to do now is I want to flip it to the folks that are in the audience to see what they are thinking about this. I’m curious about how this is all landing for everyone and what’s happening in your organization. So we’re going to put up a poll here. Put up a poll. Throw up a poll. Sounds weird. We’re going to put up a poll, and we would love to get your feedback on what best describes your hiring process today. So I know we’ve kind of boxed you into some choices, but pick the one that feels closest to you. For those of you who may be listening to this later, the question is: what best describes your hiring process today? Our options were: Evolving, maybe slightly out of breath trying to keep up. It’s mostly legacy, with a shiny layer of AI on top. We’re actively redesigning it; send snacks (favorite one). You’re right. Depending on the team, the role, and the phase of the moon. So those are your four choices. So go ahead and throw in your thoughts, and then we will share the results.

Michael Watson 17:13

But yeah, I was actually going to vote on this, and then I see down there on the bottom, “Panelists can’t vote.” So, okay, listen, I think we’re gonna land… yeah, I think it’s gonna be “It’s mostly legacy, with a shiny layer of AI on top.” But maybe… maybe I’ve been like in the basement for too long designing courses. I need to come up for air more often. But I think it’s going to be that “Send snacks.” Yes, send me snacks. Please. Coffee, too.

Rebecca Warren 17:47

Alright, let’s see where we’re at. And… oh, I love this mix. Y’all look at this. So we’ve got pretty even… the only one that folks aren’t doing as much of is actively redesigning. No snacks for the rest of you. So folks are evolving, putting that shiny layer of AI on top, or maybe at the front or at the end. And then, depends on the team, the role, the phase of the moon. So people are all over, which totally makes sense for what’s happening today. So let’s… let’s talk a little bit about that friction that we have here about where, when you introduce AI into a process, what happens. So we are calling this “the amplifier effect,” and looking at that tension. So, AI is that accelerant. It doesn’t fix broken systems, but it turns up the heat on whatever it enters. So… and it’s not, honestly, it’s not like in a slow burn kind of way. It kind of sets things on fire pretty quickly. So before we go too far into this, we’ve talked about agentic, we’ve talked about AI. We haven’t really defined it. And while there are a lot of layers to what AI is, starting with machine learning, LLMs all the way through, I think let’s just level-set on three basic buckets of what we’re looking at for AI. Because, Mike, you and I talked about this before: a lot of times people say, “We’re using AI,” when they’re actually automating processes, which there is nothing wrong with automation; it’s just not AI. So we want to be clear about what we’re trying to do and why we’re trying to do it. So let’s first talk about automation, which we look at as the rule follower. You’ve got pre-designed steps, you’ve got triggers, you’ve got workflows. It does exactly what it’s told: nothing more, nothing less. And if there isn’t a direction in there, it doesn’t get done. That’s right.

Michael Watson 19:56

If A, then B. If B, then C, exactly. If it says, “If D, then, you know, G,” then the system, when it gets to D, is going to be like, “I don’t know what to do,” right?

Rebecca Warren 20:07

Right. And if an agent is following an automation process, the agent will not be able to change it. That automation is what it is. It is a box. It’s not AI; think about it more like the structure, maybe a playbook, but it is not a player, right?

Michael Watson 20:25

So there’s so many flavors of AI, like, that was very rudimentary entry-level AI, right? Let’s talk about… yeah, yeah, go ahead. No, you go ahead. I want you to talk about Gen AI. Yes. So, right? And then real quick, in between that, you also had, like, predictive analytics, AI, that sort of thing.

Rebecca Warren 20:45

I kind of took those out because there’s so much to try to absorb, right?

Michael Watson 20:49

And think about that as, like, a Stitch Fix. I buy something, I say, “Here are the designs I like,” and then it says, “Okay, you like that, here’s more of that,” right? But what we’re talking about is a real neural network, a large language model, right? So go ahead and go back to your question again here, real quick.

Rebecca Warren 21:06

Well, so let’s think about what Gen AI is. So if we talk about automation, even talking about predictive analytics, it’s still rules-based, right? “If this happens, then we do this.” Gen AI, right, is the doer. Those are your Claudes and your ChatGPTs and your Geminis and Grok and all of those. So let’s talk about, you know, what that looks like. It’s not rules-based. It’s still absorbing information, but it produces new content based on patterns and the large amount of data that it’s able to process. So it creates, but it can’t act.

Michael Watson 21:45

It’s not fully autonomous, thinking on its own. That’s right. It’s only a result of all the information that’s been produced in the history of humankind.

Rebecca Warren 21:55

Yes, that’s right. Yes. So when we think about that, exactly, right? So it feeds a lot of things that happen in the agentic world, and it may actually help you create your automation, but it is not driving those decisions. So Mike, talk to us about agentic. So if we’ve got automation, we’ve got Gen AI creating things… where’s the doer, right? So we’ve got the creator, we’ve got the rule follower. Now we need the doer: agentic AI. Talk to me about what an agent is or where the agency comes in.

Michael Watson 22:35

Yeah, think about this: it’s just something that can take action for you. So for instance, I create an agent that would go out… and I love pizza makers, right? So I created an agent that would go out and look for Facebook Marketplace for pizza ovens and let me know when I had one, and I could see if I could buy it at a good price. So, right? We can create agents for everything, right? We create an agent now that will go out and interview your people for you. And not just that. Because here’s the thing: anyone can go out and create an agentic agent right now. You could go out and do it. You can go out and use n8n. You could use ChatGPT. Lovable, right? You can get your API key, and you connect the two, and you’ve created an agent. But there’s a difference in agents, right? And I would encourage all of you, when you’re looking at agents, look at the size of the dataset with which that agent is enriched, because that’s going to be the difference between an agent that knows your business, can conduct the types of activities that you need done, versus just a generalized one that I created on ChatGPT 10 minutes ago. There is a vast difference.

Rebecca Warren 23:55

So if we think about that, that was perfect. So if we think about automation follows the rules, like a chatbot. We’ve got Gen AI, creates the content, like a copilot. And then you’ve got agentic AI, which connects, decides, and acts. So it’s with agency. So when we talk about… when we think about AI and hiring, a lot of what’s in the market today is still sitting in those first two categories, right? It’s a lot of automation. It’s a lot of Gen AI. All are good, all help you become more productive, maybe more effective. So maybe that’s why teams feel like they’re successful with AI, because they’re working faster, but they’re not working differently. So we’ve upgraded the outputs, but we haven’t fundamentally changed how the work gets done. So, Mike, you talked about some of the things that you’re seeing. Like, what do you think is one thing… one thing that automation has made faster in recruiting? Like, I think scheduling probably is the easiest one. But what has automation helped with?

Michael Watson 24:56

It’s definitely helped with scheduling. Hmm. It’s helped with, you know… one of the workflow automations I helped one of our customers build was around employee referrals. Okay. They had a—and I love this because I had the same philosophy when I was running TA—but they had a philosophy of, “Every single referral gets an email from a recruiter and gets a screening. No matter what. I don’t care if they’re qualified. They’re not qualified. Our employee referrals are so important to us that we don’t ever want an employee coming to us and saying, ‘Hey, what happened to my referral? Did you talk to them? Did you screen them?’ So we’re going to screen everyone.” So we built in an automation that anytime someone was referred in, they automatically got sent a link to my calendar, or whatever that recruiter’s calendar was, to interview for that role. Okay, the recruiter didn’t have to look at it. The recruiter didn’t have to look at a resume. All of a sudden, a meeting pops on their calendar. You know, it got to a point where my recruiters were really just kind of showing up every day and, “Hey, what’s on my calendar today?” And that’s what the agent is loading for me, and that’s what I’m going to do. And they really enjoyed that, right? It’s like, yeah, cut out the clutter, right?

Rebecca Warren 26:16

Okay, so that is a case of maybe where recruiters were the connective tissue that they didn’t need to be. I did a webinar recently where I talked about TA processes that were over-coordinated and under-designed. That over-coordinated piece is where recruiters are doing things that they don’t need to do, right? Like thinking about the scheduling piece, the back and forth, thinking about some of the interview, you know, setups or processes, things like that. So where else do you think—and I think it lends itself right to screening—but where are recruiters still acting as the connective tissue between systems, and maybe where they don’t need to be, or they shouldn’t be? Maybe there are different things that they should be doing?

Michael Watson 27:05

One thing—and I would love some comments in the chat from… oh yeah, throw things in the chat. You have some things in here. But one thing that I see… this was a giant time eater for me, was after I interviewed you, Rebecca, taking my notes, coordinating my notes, doing my write-up to the hiring manager, right? That can all be automated now. So once again, I make sure that my assistant, my recruiting assistant… and we just came out with our newest recruiting assistant to go along with our agentic agent as well. But I make sure I have my assistant on, so it’s taking notes. So once we’re done with that conversation… I mean, I remember three or four years ago, when you’d jump into a meeting and people were like, “AI is on, it’s taking notes. I don’t want that. Can you turn it off, please?” Oh yeah.

Rebecca Warren 27:54

Folks didn’t even want to have cameras on, much less getting out…

Michael Watson 27:56

Or, “I don’t want to jump into a meeting unless it has an AI note-taker, because I’m going to forget all these things.” No, right? So it’s shifting, right? And so, yeah, I think that’s a big one where I think some recruiters are probably doing it. I think the ones that are probably really efficient and are knocking out, you know, 2x, 3x their goal on positions filled are probably doing something like this. I think the rest probably aren’t.

Rebecca Warren 28:28

And folks, I think, to your point, folks are a little bit scared of it sometimes. You know, it’s the bigger thing about, “Is it going to take my job?” And we’ll leave that for a minute because that’s a whole other ball of yarn. But I think what folks get scared by, which I think is fair, is that it’s not that AI isn’t working, right? We talked about that. It’s doing what it was intended to do. It’s exposing what was already there. So if your process was clunky to begin with, AI isn’t going to fix it. It’s going to show those gaps even more. So as you talked about throwing things in the chat, let’s get some audience interaction.

Michael Watson 29:06

I’ll give you a real-life example where I would love for you to do that and expose the gap. Yeah, right. And it was the classic, to all my TA friends on this call: you’ve gotten this, “Hey, how come you’re not filling positions faster? How come your team’s not filling positions faster?” right? Once we had…

Rebecca Warren 29:22

Because they’re just sitting around eating chocolate and watching TV, like… right? Like…

Michael Watson 29:25

I’m talking to my recruiters: “We’re screening people a day after they apply. We’re getting people to…” So I had to get the metrics and track. And one of the nice things about AI is I now have metrics, right?

Rebecca Warren 29:37

Huge, right? The reporting was the thing that always made me want to stab my eyes out. Now it’s so much easier to get.

Michael Watson 29:43

When I looked at the data, we had no problem screening people once they applied, getting people to the hiring manager, getting them to the interview. Once we got offers to finance, it took finance 11 days to approve an offer. Finance was your bottleneck. But the business doesn’t know that. The business just knows that their position’s not getting filled and “what’s going on, what’s right?” So they’re coming beating us up as the business owner for that, right? And so I did the analysis, and had the data, and I saw which stage of the process that these things were getting hung up in. And I went to my CHRO, and I went to the CFO, and I said, “Hey, what’s going on here? Where can we improve things?” Yeah. Well, think about how much faster… “That’s right, the finances were approved. It’s within range. Why is it taking 11 days?”

Rebecca Warren 30:41

And how much faster is that exposed when you put AI into that process, right? You had to go through and dig through the data, figure out what was happening. AI is going to show that. Good or bad or ugly, right? It’s going to come out. So… so let’s throw this to the audience. So let’s… let’s talk about what has… so we’re going to put up another poll. And what has AI or automation improved the most in your process so far? I think there’s a lot of ways that maybe lives are getting a little bit easier, like I said for me, that data insights, that reporting, that stuff that was so manual before. But your options are: speed of tasks, candidate communication, data insights, flow of work, or, honestly, not much yet. And we also understand that there are some very regulated, risk-averse industries or companies where they’re just like, “Absolutely not.” I think you can’t do that forever. Maybe you can do that for now and feel like you’re going to be okay, but it’s… you’re going to have to think about that. But it may be right now, there’s a lot of organizations that are just really risk-averse and scared.

Michael Watson 31:54

So Rebecca, I’m gonna try and guess the results again here real quick. “Honestly, not so much.” I don’t think we’ll have much there. I think everyone’s trying to absorb everything right now. For me, I think it would be “speed of tasks” or “data insights,” because those are the two examples I gave, like real-life examples that I gave you. I think “communication” is where people are still hesitant.

Rebecca Warren 32:16

I think it’s not good. Like I talked about people who are in the job search, and that communication is not flowing. At least, at least not at the very upfront, it’s not flowing. I had a friend of mine who applied for 318 jobs over the course of two years, and she was ghosted by over 70%: no response. So I hope, I hope, if “candidate communication” is selected, I hope it’s from folks who are doing this right from the onset, right from the minute that somebody applies all the way through the process.

Michael Watson 32:50

Yeah, and if you did select “candidate communication,” drop into the chat what exactly. I would love to hear. Alright, let’s see. Drumroll.

Rebecca Warren 33:02

I think you’re… look at that: “Honestly, not much” was not chosen. That’s great. “Speed of tasks” some are doing. “Candidate communication,” please tell us what you’re doing. Chat us separately if you don’t want to put it in the chat; throw something on LinkedIn. We want to know what that looks like. “Data insights,” “flow of work”… okay, okay. I think that all makes sense, and I think what we’re learning is that automation and AI can certainly help, but it’s not necessarily fixing what’s underneath. We’re helping a process that probably needs to be completely evaluated and maybe redesigned. So let’s talk about what actually needs to change. So when we think about what this looks like going from layering to re-architecting—which, Mike, we’ve talked about this: it’s really hard, and it’s painful, and it feels personal, and it feels like the ground is being ripped away when we say the way that we’ve been doing it for 20-some years is no longer the way. It’s a big shift. So moving from stacking tools to redesigning the flow of work. And in everything that I’ve learned and seen and heard, this is not about a bottom-up or top-down process. This is a lattice; things are changing. So Mike, let’s talk a little bit about that. If we could wipe the slate clean and we go into a company that has amnesia about everything that’s been done over the last 20 years, we could wipe that slate clean, how do you think we should design hiring today? What’s important? Where do we… where do we land if we could blow it all up and start over?

Michael Watson 34:57

Now, this is me, right? I’m going to start with a nice user interface because I think candidates, 70, 80%, basically know where they want to work before they even apply. Okay, recruiter branding, marketing, you know, our career page, I think, makes it easy for that. So I’m going to make sure I have that very easy to apply, right? I don’t want cumbersome. I don’t want link-offs. I don’t want you to go into multiple different systems and multiple different URLs from an internal process. Or, you know, if I’m the CHRO and I’m rebuilding this again, I would go with agentic first. Like agentic…

Rebecca Warren 35:42

The whole entire stream, you think about like a swarm, or an entire…

Michael Watson 35:47

Quick example, right? I used to be with an HRIS system that was cloud-based, and our competitive advantage was we used to go against the other HRIS systems and say, “Hey, yours is in a database somewhere. It’s not cloud-based,” right? And so now we know that that dream has disappeared, right? There’s no more efficiency left in the cloud. SaaS is somewhat dead, right? You see the SaaS companies getting beat up, and the reason for that is it’s very, very hard to take a legacy software application and layer in agentic AI on top of it. So that same competitive advantage I had 10 years ago when I was at the HRIS system and I was saying, “Hey, we’re cloud and we’re SaaS,” the companies that are going to win… if I was building this again, I wouldn’t use a SaaS or a cloud company. I would use an agentic-first, AI-first company because it’s going to have everything I need to build on top of it. Right? It’s the old debate: VHS versus Beta, right? Record, 8-track, cassette.

Rebecca Warren 36:55

It’s… you might be dating yourself here, Mike, and I’m here for it, because I’m in it with you.

Michael Watson 37:04

What I’m saying. So I would make sure… if I’m doing this over today, I would make sure that all of my vendors… and listen, I’m doing this right now when I’m talking to my SMEs, right, when I’m talking to my LMS (learning management system)…

Rebecca Warren 37:19

You’re talking to your texting… okay, I do that talking too.

Michael Watson 37:24

You know, I’m talking to my LMS providers, and I’m talking to all my providers, right? I’m asking them about, like, “What’s your AI vision? Where are you folks going? Yeah, yeah. What are you doing with agentic?” You know, one of the things I would love to personally find, talking about AI first, is, right? I run training. So I can right now build a bot that will present a deck to you. What I’m looking for… the vendor I’m now looking for is who can build a system that can not just present the deck, but when Cathy or Stephanie has a question in chat, that agentic bot automatically pauses the chat and says, “Hey, Stephanie, let me pause here real quick and answer your question.” Yeah. We should all be doing that. If you’re not doing that, you need to start immediately with every single vendor you’re talking to.

Rebecca Warren 38:19

So Mike, to jump on that… you know, Ashu has talked about that, and I wrote an article about how we need to shift from being a SaaS organization, right, software as a service. We’re actually working at Eightfold to shift that to become service with software. So what does that service look like? Right? You talked about the UI, you talked about the process. What does that service look like? And then we’re going to put software in there to help you do the things you need to do, which is agentic AI, which is the other processes. So flipping that… even that thought, which is what you just said: don’t look for a SaaS company. Look for service with software. Service with those tools and those pieces built in, right? And once again, if…

Michael Watson 39:05

I’m redesigning today, I’m looking for an organization that will do all of that upfront for me and then present to me these, in essence, talent cards that I can look at and I can assess, and I can say… I can go back and watch the interview. I can, you know, watch the strings. Then I can make the decision. Human in the loop, right? Yeah, really.

Rebecca Warren 39:27

I talk about human in the lead. You want…

Michael Watson 39:29

The human making these decisions for you, right? So, yeah, I can get a system that does the interviewing, does the note-taking, does the writing for me, and then it presents it in a very easy-to-follow, read format that I can then take those… forward them off to my managers. I can reach out to them for an initial conversation myself. Yeah, think about this: it’s like the old network operation centers, like a big NOC with a bunch of monitors over there. I envision a recruiter of the future walking into, like, a network operation center with multiple screens, multiple monitors: “Here’s my requisitions in flight. Here’s my requisitions being approved. Here’s my offers in flight. Here’s where the AI agents are helping me and acting.” That’s right. That’s right. That’s right.

Rebecca Warren 40:16

Yes, yes. I love that. I love that. Okay, so I think that’s maybe where the tension is coming from, if we needed to call it something. You know, you talked about, once you figured out where the bottleneck was, where the friction was in your hiring process, it was taking time with finance, right? You take that out and you see what a less-friction process… “less friction,” I don’t think that’s a word. If you think about a process with less fiction… friction in it, oh my gosh, it makes a huge difference. So I think that flip side is when you remove that friction, you start to see what that process could look like. And I think that’s where maybe some of that tension is coming in too, because we’re designing for control in a world that’s moving towards adaptability. So let me go ahead and… we’re gonna… we have one more.

Michael Watson 41:08

I always think about this…

Rebecca Warren 41:16

Okay, go ahead, and then we’ll put up the poll.

Michael Watson 41:19

Yeah, our ATS systems were built for one thing. It was for compliance. Oh yeah. It was to track people, for sure. It was never to make your recruiters’ jobs easier. No, no. That’s what we’re trying to do. We’re trying to say, “Okay, you have a compliance piece here that we need to follow. But how do we follow that and also make your recruiters’ lives easier?”

Rebecca Warren 41:32

Yes, huge difference. That’s right, ATSs were the container to make sure that we were compliant, that we weren’t calling the same person, that we had feedback notes for OFCCP, like all of those things all in there. We had those too. Okay, so let’s get a poll out to our folks talking about, where would you most want AI to take action—so to be agentic, not just assist? So we’re not talking about automation. We’re not talking about a copilot. We’re talking about an agent that would be able to come into the process, do the things, make those decisions, and then get to that end… that end result. So is it screening and shortlisting, scheduling and coordination? Is it conducting structured interviews? Is it synthesizing insights, or is it somewhere else? So out of those five, like, where are you… where do you think? And maybe thinking about where our customers are, like, where are folks landing where they say, “Gosh, we love this, and we would love to have that”?

Michael Watson 42:45

I feel like the old Carnac the Great here, trying to predict the future. I know the magic. So screening, shortlisting… yeah, I definitely think. I think the one I want to talk about here: conducting structured interviews. I don’t know if we spent a lot of time talking about this; I’m not sure the broader audience has thought about this. But I think that is probably the biggest one on here, but the one that might get the least votes. So I’m interested to see, and I could talk about why that’s so important later, or a slide or two. But yeah, I think all of them are important. I don’t think we’ll see “somewhere else.” We might see one or two on there, but I think it’s going to be primarily screening, shortlisting.

Rebecca Warren 43:30

These are the ones that tend to hit us in the face most of the time, right? When you’re thinking about a TA process, the things that are the most visible. And I do want to… when we see the results, I do want to talk about screening and shortlisting, regardless of how the results come out. So, alright, where are we seeing the answers to this one? Okay, there is a “somewhere else” in there, but interesting: you mentioned “conducting structured interviews” that…

Michael Watson 44:05

I said not many people that picked it, but it’s an important one. So, yeah, almost like I had… folks, I promise you, I didn’t have advanced… I can’t see the results coming in. I promise you. I’ve been doing this a long time.

Rebecca Warren 44:21

Well, so… so let’s talk about that screening piece, right? So before we flip to our next section, going from theory to reality, let’s talk about that screening piece and what that actually does. Now, I know that there’s a lot of hesitation from folks saying, “I don’t want to be screened by a robot,” or “I don’t want AI to be all up in my head.” But we have seen a ton of value that comes out of that, not just from a responsive and interactive process. And let me say this first: I was really reluctant. I was like, “Oh no, that’s terrible. I’m going to hate that.” And I actually don’t hate it. And I’ve used our AI Interviewer several times, even last year when we first came out with it, and I was like, “Oh, this is kind of cool.” And it’s continued to get better. So the idea of a responsive, reactive agent having a chat with you… you know, folks were nervous about it, but it actually really does make a difference. And if you think about if you were trying to solve some of that upfront process, right? Like, you talked about everybody who is an employee referral got an interview. What if everybody who applied to your role actually got an interview, regardless if they were the right fit? Think about what kind of information you would have in your talent pool if everybody got one. Even if it was just a 10-minute or 15-minute, “Just tell me what’s the most important thing about you.” People can do it on their own timeframe. They can do it when they want. Like, I don’t know… to me, that just feels like a really easy way to change the process and have people feel seen and heard.

Michael Watson 46:03

Rebecca, I’ll make a kind of a funny comment here. If we have any lawyers on this call, you’ll like this. I look at some of the lawsuits that come out in the HR space, and many times it’s, “You didn’t interview me. I applied for the role. I thought I was qualified. You didn’t interview me.” When I see these—and this has been the last six months, right—every time I see one of these coming out, I can’t help but tell myself, like, “Man, if they would have had an AI interviewer there at that first stage, interviewing everyone, it deflates that argument right out the gate.” Everyone gets a chance here, right? It’s powerful. It’s powerful. Now, I understand there are organizations like… you have hundreds of thousands of people applying a day. I got you. There’s scale to this. So maybe you calibrate, and you select all your top-ranking ones and go that way. But yeah, it’s…

Rebecca Warren 46:59

I don’t know. There’s… there’s something there. I think there’s a nugget. Okay, so I’m watching the time here. We’re going to flip to what this looks like in reality. So how are we actually thinking about redesigning work in this agentic space?

Michael Watson 47:17

I apologize to you. I do want to make a comment on that structured interview real quick. Okay, because… because folks, and this is the beautiful thing about it, too. And I talked about this… I wrote this in a white paper. If you are really interested in measuring your quality of hires, you must have some type of structured interview process in place, and you got to make sure that that is adhered to 100% of the time. Which means the same questions you ask your candidates on January 1, 2026, are the same questions you’re asking your candidates on December 31, 2026. And only then can I go back and look at the quality of my hires—who’s reached, you know, exceeding expectations, all of that stuff. It starts with that structured interview. So I know we didn’t talk about that, and I said it was important. So for those of you folks out there that are like, “Oh gosh, I’ve been trying to find a way to measure my quality of hires,” it starts with structuring your interviews.

Rebecca Warren 48:16

No, it’s perfect. It’s perfect because it leads into something that was on my mind, too, about the idea of, you know, if you have a system… like, I think about when I was doing interviews, too, like there’d be times at the end of the day I’m like, “It’s not about the candidate. It wasn’t personal, but it was like, I’ve had a gazillion things going on. I don’t really want to do this interview, but it’s on my schedule, and so maybe I wasn’t 100% there.” Maybe in the morning I’m better or worse or whatever, right? There’s always these variables. So that thing about fair, and consistent, and documented—all of those pieces, right? When you think about that going through an agent, it really sets that up differently because you know that you’re getting that consistency. But let me… let me ask you a question this way. So if you had to say, like, what’s fundamentally different? And maybe I’ve already said it; I don’t know. What’s fundamentally different about AI conducting interviews via the traditional screening model that I talked about? Maybe I’m in the car with my kids taking them to an event or whatever. Like, what’s… what do you think is fundamentally different about AI being able to conduct those interviews versus a traditional screening process? Well, we already covered it all.

Michael Watson 49:35

No, no. I think for the average person that works nine to five, you won’t see much difference. I think where this thing really helps… you know, I ran IBM’s Western Digital… it was a subset of IBM, our disk manufacturing. We ran three different shifts. We ran a midnight shift, a swing shift… where I see AI Interviewer coming in is on that person that gets off at 11 PM and doesn’t go to sleep till 3 AM. So they’re not up during regular interview hours. But with AI Interviewer, at 1:30 AM, at 2 AM in the morning at their kitchen table when they’re, you know, having dinner and they’re getting ready to go to bed at night because that’s what works with their shift, they can now take that interview. Like, yeah, I look at this as completely leveling the playing field. Like, I’m such an advocate for leveling the playing field for people that don’t have degrees, for working-class folks. Like, let’s make this easier. Let’s tap into what skills they have, and then let’s match them to opportunities and let’s hire them.

Rebecca Warren 50:39

Yeah. Well, and I think that’s what we’ve been talking about, right? That is… this is my… one of my favorite quotes here, right? Like, “Agentic AI should shrink the administration so humans can expand the impact.” I love that idea of, how do we take out the work that can be done without human intervention and let the humans do the work—that orchestrator, that strategy, those pieces of saying, “Where do we need to make sure that a human is doing this work, that is evaluating, that is looking at things, that is being agile?” All of those pieces, I think that’s really where we see the biggest difference. And if you add in… we had…

Michael Watson 51:24

Someone actually made a comment. John said, “Interviews across time zones. Agentic answers.” Yeah, exactly, right. John, it levels that playing field. So I don’t care if you’re in Australia, I don’t care if you’re in San Francisco. I don’t care where you’re at. You can be in Prestonsburg, Kentucky, right?

Rebecca Warren 51:42

Yeah. And I think… let’s just say this here, and this is just a statement before we wrap it up. Because I know I’m watching the time. I do think there is a shift coming. Right? Like, let’s be honest, the shift is probably already here. We hear people saying, “We’re downsizing.” Oracle just talked about, I don’t know, like 20,000 people that got emailed saying their positions are being eliminated because AI is coming in. Like, I think we have to be real to say that AI is changing the game. But I don’t think it needs to change how the TA function operates if we are prepared to say, “We understand AI and we are going to now be the ones that own it, lead it, drive it.” So yes, some roles are going to be compressed, some are going to be changed, some are going to be eliminated. But I think the work is still there. It’s just at a different level. So Mike, you had brought that up earlier, and I did want to say we just need to address that before we close it out. Anything you want to add on that before we hit our final?

Michael Watson 52:45

No, I think I’ll share with you folks the same thing I shared with my team back in 2018 when we implemented AI in HR. “Folks, I see this trend coming, and I don’t see it waning off.” And now we know for sure it’s here. It’s not going anywhere. So if you haven’t yet, lean in. I honestly think it’s been the biggest blessing in my career, was coming to work here at Eightfold, right? Because I was already… I would agree with you, right? I was already doing things with AI behind the scenes. And when I came here, it just supercharged it. So I agree, and I’m not bragging, but I am: even though I’m not an engineer here at an AI company, I am a person that our product teams come to, our CS teams come to, for my AI expertise, because I am curious. I build bots. I go to n8n. I take my API key from ChatGPT, and I connect it and see what I can do. And I do all this for fun. That’s sick, right? But, you know, I would encourage all of you to be that way. If you all lean into it this way, you won’t have to worry about AI taking your job, because you’re going to be the one person in your organization that understands exactly how it works. They can’t get rid of you. So lean in, lean in.

Rebecca Warren 54:13

I agree. And Mike, you know, when I first had my resume floated over to Eightfold, I thought it was for a TA position. And I think you thought your role was going to be a while… and they were like…

Michael Watson 54:26

TA, and it wasn’t! Like, “Yeah, we think you’re a better fit for something else.”

Rebecca Warren 54:31

To create Customer Success. You’re like, “What?” And so when I interviewed and my resume came over, I was like, “I have no idea. I don’t know tech. I don’t know anything about customer success.” But the process of the interview process of putting together a deck and doing a presentation, it was about the skills that I was able to transfer from TA on over to Customer Success. And I would say to Mike—so Mike led the team, and so I reported to Mike—and I would say to Mike, I would say, especially at the beginning, almost weekly, like, “Mike, I suck at this. I’m failing at this. I did this wrong. I did that.” And Mike would always come back and like, “Nope, we’re all learning together. It’s okay.” And I would say I’m uncomfortable, and I would tell people that, and it wasn’t a bad thing. But I’m like, “I’m uncomfortable every day because there’s so much stuff I don’t know.” But the folks at Eightfold—Mike, the rest of the folks around us—leaned in and said, “Here’s how we’re going to do it. Here’s how we work through it.” Like, had I not had that “thrown into the fire” moment because somebody took a risk to say, “Let’s bring her in,” I would not be where I’m at today. And the same thing about moving from Customer Success to Marketing: it was a brand-new team. We built it. Jason Serrato made it up and said, “Hey, we’re going to pull in people who haven’t done this before, but we’re going to build it.” Those are those things. Stay curious, stay humble, stay interested, stay uncomfortable. That’s where you’re going to find that shift, I think, from hiring falling behind to AI hitting that fast forward. We need to jump on that “AI hitting fast forward” idea to think about what that looks like in the future. Okay, so my last thought before I send it back on over to Kathy, is… here’s my final spicy truth, or my line: Organizations that figure this out won’t have the most tools. They’ll have the fewest gaps between how hiring should work and how it actually does. Mike, closing thoughts?

Michael Watson 56:26

Yeah, I think that’s right.

Rebecca Warren 56:30

That was pretty… that might be the first time that we both said, “Yeah.”

Michael Watson 56:34

So final thoughts: lean in. Lean in. Start small. Don’t try to boil the ocean. Pick a team or a group that’s curious, that’s going to experiment with things at night. You’re not asking people to do anything out of the ordinary. But you know, I’m always on my laptop at 9:30, 10 o’clock at night playing around. Yeah. And think if you really want to get good at this, understand that that’s what it takes. It takes that intellectual curiosity.

Rebecca Warren 57:07

That’s right. So reach out to us if you have questions. Find us on LinkedIn. Find us on our pages. You’ll see us on our website. Check out the Talent Table, which is a webinar series that I run. We’d love to have more chats with you. Appreciate you all, and we’re going to send it back to Cathy to close us out.

HR.com Moderator 57:27

Thank you all. Thank you, Rebecca and Michael, for this very informative webcast today. Stephanie said in the chat she wants you to send her that quote. Got it. Got it. Alright, so just a final reminder to the audience to just fill out a quick feedback survey that’s going to open on a new tab on your screen. It only takes a few minutes, and your valuable input is key to improving the quality of our webcasts. We hope you can join us at the next session. It’s going to be starting in just a few minutes. Click on over to the HR.com event auditorium. We hope to see you there, and enjoy the rest of the event. You.