Pod Insights The Chad & Cheese Podcast The Human Hiring Touch with Terry Baker About this episode In this episode of The Chad & Cheese Podcast, Terry Baker, CEO at Daxtra, dishes on how AI is shaking up recruitment like a caffeinated barista on a Monday morning. With 23 years of AI street cred, he insists AI’s only as good as the data it’s fed—garbage in, garbage out, folks—while preaching the gospel of keeping hiring human, because no one wants a robot therapist when the job hunt gets rough. The convo takes a wild turn through resume validation (think CSI: LinkedIn Edition), with Baker warning that AI-generated CVs are sneakier than a cat burglar in a yarn store. From automating soul-crushing high-volume gigs to spotting skill gaps in employees who still think “Excel” is a workout move, Baker’s got big plans—capped with a future where AI’s so slick, it might just write your performance review and roast you at the same time. Chapters 00:00 - Introduction to Terry Baker and Daxtra 02:53 - The Evolution of Daxtra's Technology 05:53 - Engagement and Humanization in Recruitment 09:14 - AI's Role in Recruitment and Performance Reviews 11:54 - Validation and Verification in Hiring Processes 15:08 - Navigating DEI and Meritocracy in Hiring 17:51 - The Future of AI in Recruitment 20:59 - Closing Thoughts and Future Directions
Transcript 0:00 Speaker 00 Dude, it turns out that Winston Smart Recruiter's first AI hiring sidekick is taking all the grunt work out of Recruiter's Day. Things like candidate screening, scheduling, all the admin garbage no one actually wants to touch. Winston can do it all.
0:17 Speaker 01 Man, you know I'm a fan of using new tech that takes the boring repetitive crap out of a Recruiter's Day and allows them to have more time to connect with hiring managers and candidates.
0:29 Speaker 00 Yep, Winston is doing the heavy lifting while recruiters kick back and do the actual hiring stuff. You know, like talking to people.
0:37 Speaker 01 I really dig that Smart Recruiters is evolving out of the legacy ATS model and embracing cleaner processes using AI. With most ATS providers, it's like, hey, 2003 called and it wants its 30-minute application process back.
0:53 Speaker 00 No doubt. It sounds like Rebecca and the team are using Winston to push Smart Recruiters into the future while the ATS vendors are basically on an AOL dial-up modem with a dashboard.
1:05 Speaker 01 With Smart Recruiters, Winston does the grind and the recruiters get the glory.
1:11 Speaker 00 For more information on superhuman hiring with Winston, visit SmartRecruiters.com. That's SmartRecruiters.com.
1:19 Speaker 01 Start using Winston today, or I guess you can stick with...
1:25 Speaker 02 Now we return you to our regular scheduled program in progress.
1:29 Speaker 03 Hides your kids, lock the doors. You're listening to HR's most dangerous podcast. Chad Soash and Joel Cheeseman are here to punch the recruiting industry right where it hurts. Complete with breaking news, brash opinion, and loads of snark. Buckle up, boys and girls. It's time for the Chad and Cheese Podcast.
1:49 Speaker 00 Ohhhhh, yeah. This is the Chad and Cheese Podcast. I'm your co-host Joel Cheeseman. Joined as always, Chad Soash is in the house as we welcome... Terry Baker, CEO at Daxtra. Terry, welcome back to HR's Most Dangerous Podcast.
2:12 Speaker 04 It's good to be back with you guys. It's been a while.
2:15 Speaker 01 The first time, was the first time in Ireland? Was that the first time you were on the show?
2:18 Speaker 04 No.
2:19 Speaker 01 When we were at the Guinness? It wasn't at the warehouse? So that was the second time? Okay.
2:22 Speaker 04 I think so.
2:23 Speaker 00 I'm glad you added "on the show" to that was the first time at the Guinness fact. Yeah, on the show. Yes.
2:28 Speaker 01 On the show.
2:29 Speaker 00 Terry, you know, we're adding fans all the time. So there are people that don't know who you are, haven't heard you on the, on the show. So give them, give them kind of the elevator pitch on you and about Daxtra.
2:38 Speaker 04 Sure. I've been in HR tech for the hard to say, but 30 years. Formerly at Panda Logic, Real Match sold that company for 200 million. Got a great exit. Thought I was done. Went three months on the beach.
2:53 Speaker 01 Give me that money.
2:57 Speaker 04 And, you know, I communicated with Chad when I was on the beach and I was like, this is really fun. And then one month went by, two months went by, it's kind of got a little boring. And so I started doing some private equity capital work and a company called Stratum pulled me in on a due diligence they were doing with Daxtra Technologies. So muscle burrow backed company that's been in AI for 23 years, which is amazing. And so they pulled me out of retirement and got me back and I've been enjoying the ride ever since. I haven't been to the beach in a year.
3:35 Speaker 00 I was gonna say, is the yacht getting gassed up? Is that why you're calling us from a room? You're usually, you know, you usually look like a Duran Duran video. Yachts and like, you know.
3:46 Speaker 01 He's hungry like the wolf.
3:47 Speaker 00 Cocktail shrimp and yeah. So it's, it's nice of you to tone it down for the show, Terry.
3:52 Speaker 04 We did have a good day on a yacht in Newport Beach. That was fun.
3:57 Speaker 01 I know. It's good times. Good times. Hoping for more of those. Hoping for more of those.
4:01 Speaker 04 Well, let me tell you a little bit about Daxtra. They've been doing natural language processing, machine learning, parsing resumes, job descriptions for 23 years. And I think that Data that HR tech is based on, you look at any ATS system, if they don't have a good parser, they're not getting good data. So AI is only as good as the data you generate. And we've been generating really good data for quite a long time. And now we've moved kind of down funnel. Now we don't just parse resumes, job descriptions, but we do grading, scoring, ranking on scaling data, enable recruiters to see who the most qualified applicants are in just seconds.
4:45 Speaker 04 And then we do the engagement part. So if you've found the most qualified candidate, you don't want to send them to a chatbot.
4:53 Speaker 01 You pivot.
4:55 Speaker 04 You want to engage them directly. Pivot, that's a good word. We now call it Dexter Engage. And that engagement helps solve the black hole, getting conversion of candidates, particularly the most qualified, quickly, effectively, and with little effort for the recruiters. So that's what we do.
5:14 Speaker 01 So let's talk to this real quick. So you get a job description, and tell me where I'm going off the rails here. You get a job description, you parse it, you rip it apart, you contextualize it, you know exactly what the requirements are. Then you jump into the databases, whatever those databases are, external, internal, yada, yada. Then you do the matching. At the matching point, you hit a certain threshold, whatever the company wants that to be. It's a 90% match or what have you. And then you engage those individuals to be able to either get more information and/or just to apply. What happens after that? Is that, first and foremost, is that a, is that a good picture? Okay.
5:49 Speaker 04 That is a good picture. There's one little piece that we provide during that process, and it's called the DAC Agent. A lot of companies are getting into agentic, you know, agents.
6:00 Speaker 01 Agents, cheese men.
6:02 Speaker 04 Which are so much smarter than bots, right? They learn, they take actions, they make recommendations.
6:09 Speaker 01 They're specialized.
6:11 Speaker 04 Yeah, when a recruiter runs a search, we're learning from that specific recruiter's work log and building an agent that is making recommendations and then pushing the engagement.
6:24 Speaker 04 Directly out. So the recruiter doesn't even have to worry about it. Everybody that's applying gets a campaigned response, but the most qualified applicants, they get to talk to a human, which is what they prefer, right? Because it's what enables the hiring manager and the recruiter to humanize the engagement with the most qualified applicants.
6:45 Speaker 00 Taran, I know you, I know you like baseball. You're a big Padres fan. So I'm going to start out with a nice little, little lob for you to knock out of the park. Okay, ready? What? And you get this as a, as a sponsor of the Chen She's podcast as well.
6:57 Speaker 01 Yeah, one.
6:58 Speaker 00 Who are some companies leveraging Daxtra technology?
7:01 Speaker 04 Uh, just small little companies like, I'm going to play Manny Machado here. Hit it out. So little companies like Google, Amazon, Apple, Meta, Ernst & Young. We have a lot of people that have taken our technology and ingrained it into their recruiting platform.
7:20 Speaker 00 Yep.
7:21 Speaker 01 But here's the thing, cause you've got big tech companies like that, right? And, and some of them like Google have come into our space and they exited. They got the fuck out, right? So I mean, that, I think that says a lot, first and foremost, from the, the specialty, right? The focus in this industry is not easy to be able to not just parsing, but we're talking about matching. There's so many different aspects of what we do in our industry that even the big boys can't do. So I mean, I think that says a lot and we don't have a lot of parsing, matching types of tech in this space. Now, quick question around that because of LLMs and because of agentic, do you think that's going to change?
8:03 Speaker 01 Do you think that it's going to be easier for some companies to be able to get massive amounts of information, train those LLMs and then start parsing, matching on their own?
8:12 Speaker 04 Yeah, I think big companies are going to produce their own LLMs for specific use cases because every LLM is trained differently and they're not necessarily trained for HR tech, right? For qualification of candidates. So to some extent, we've seen parsing become a little bit of a commodity because I can take a job. Description and I can go to chat G P T and dump it in and look at what's coming out of it. But if I have to do, you know, 10,000 resumes a month, it's just not going to work. So we're focusing on the high volume hiring, highly complex jobs where you need certifications. You need not only hard skill qualification, but soft skill qualification.
8:56 Speaker 04 And that becomes really tough to do. And so we've doubled down on parsing. We're actually releasing a new parser next month that will be vector based and will extract even more data out of a job description and automatically match it to a job posting. So you get that accuracy.
9:16 Speaker 00 Let's let's touch on, Terry, I know that this is something near and dear to your heart. Sort of the the humanizing of A I through, you know, technology. And you we've all been in at this for quite a while. And, you know, the the black hole candidates don't hear back, ghosting. Like, all these things have have been a real deterrent to the job seeking experience. You're really high on how technology is humanizing the whole process. Can you talk about that a little bit more?
9:45 Speaker 04 Yeah, I think recruiters are a little bit scared right now that A I is going to replace them. And I don't think A I is going to replace humans. Humans will with A I replace humans without A I. So you have to have a recruiter that knows how to utilize A I to take care of the right tasks. But we promote the humanization of hiring through A I and automation because we want recruiters to focus on what they do best. And what they do best is selling candidates and selling hiring managers on the right candidates. All the other tasks of, you know, looking at qualifications of resumes, we simplify that. We grade and score and rank them to do engagement. We automatically run campaigns with S M S and WhatsApp.
10:30 Speaker 04 So the team based approach, so everybody within the organization can see the data that is coming in. And then data gets captured. One of the things that we do is ensure data integrity within the A T. So we we don't run without an ATS. We add value on top of ATSs and ensure that all that data, you know, I I came from programmatic job advertising. And so many candidates get stuck in recruiters' email boxes, right? They run a search, they run back the data, they look at the data, they pick one or two candidates and put them in the ATS. We automate that whole process. So everything goes in the ATS,
11:12 Speaker 04 gets updated, gets validated. We can even do data validation, right? Because in today's world, so many applicants are going to ChatGPT, dropping in a job description, job posting, pulling out the data and putting it in their resume. We're seeing resumes balloon because of the use of large language models. So how do you do validation of that for a recruiter? Who may not know that this candidate, they're ranking really good because they're meeting all the requirements from the job description because they got that from ChatGPT. So how do you validate that? You do that through third-party sources. You can put a notification, this resume was generated by ChatGPT,
11:55 Speaker 04 or it's got a lot of large language model content in it. You need to validate this through third-party databases.
12:03 Speaker 00 So the policeman, the validators, if you will, is is technology. So so Google kind of knows what content is being created by AI, similar to resumes being produced via AI. Does that turn into a game of whack-a-mole? I mean, are we just into for a future that the lazy applies of the world just get better about hiding whether they're AI or not? And then our tools and the policeman have to get better at policing that? It just keeps going on and on? Or is there an end in sight to that dilemma?
12:30 Speaker 04 Well, I think the end in sight is validation through third-party sites. And if you can do that in real time, which we do, then the recruiter knows this what was in a resume wasn't in their LinkedIn profile or wasn't in their HubSpot resume.
12:46 Speaker 00 Okay, so it's much different than a Google looking at text and saying that's that's AI-produced. You're talking about cross-referencing something like LinkedIn. To verify that. That's fascinating. Okay.
12:57 Speaker 01 And if it's AI produce, who gives a shit as long as it's true, right? The validation part is what Terry's talking about. So that validation part. And then also, I don't understand why we're not moving more toward validation screening testing, right? Like coding tests and those types of things for for many different types of positions. And I know that you guys are really focused on being able to provide kind of like those technologies and those different pieces to help companies get to that validation point. Are you guys looking at, or do you have partners where you are currently doing those kind of like coding tests, screening, those types of things to not just validate what you're seeing on their LinkedIn and other points of reference, but also actually what they can do?
13:42 Speaker 04 Yeah, that's where the DAX agent comes in. It can look at a resume, look at a job description, say where the gap is on skills. You could say for engagement with this candidate, ask them the following three questions, or give them a test on this particular, you know, criteria.
13:59 Speaker 01 The agent has testing capabilities as well? I mean, like creating tests?
14:03 Speaker 04 Eventually we'll get there. Today it's doing recommendations, but yeah, we'll get there.
14:08 Speaker 01 Okay, okay.
14:09 Speaker 00 Is this exclusive to Daxtra? Are you guys like third party? Is there a solution out there that will verify resumes that everyone can plug into? Or is this exclusive for you?
14:18 Speaker 04 No, we're using a company called Pitch Me.
14:21 Speaker 00 Okay.
14:21 Speaker 04 And Pitch Me does real time validation. We've used PeopleData Labs. We've used LinkedIn. We have a product called Magnet. You can, a recruiter can go directly and pull down a profile right off of LinkedIn and put it right into the ATS and compare.
14:39 Speaker 01 Meet Dave. Dave's resume says he's customer centric, a point of sale systems expert and a people leader. Reality check, Dave's resume was written by AI and AI applied to thousands of jobs while Dave was on his couch binging episodes of Jackass all day.
14:59 Speaker 01 How can you insert trust back into your hiring process? One word, Tatio. Tatio cuts through the BS by using real world job simulations, testing candidates so you can actually see how they perform the job before you waste a single minute interviewing or onboarding a fake. Insert trust back into your hiring process with Tatio. Check out Tatio today at T A T I O dot I O. That's T A T I O dot I O. So when you're taking a look at the landscape that's out there today, because there's so much that's happening. There are a lot of legacy companies and some could look at Daxter as a legacy company. The only, the only difference is you guys have been doing AI the entire time, right?
15:42 Speaker 04 So you're a little bit more fluid.
15:44 Speaker 01 It's a little bit easier for you guys to move. But when you take a look at the landscape today, what do you see happening with some of those legacy core platforms? Like the applicant tracking systems? What kind of disruption do you see happening?
15:57 Speaker 04 Well, I think everybody is looking at engineic agents to build value into the recruiting process. So if you're an ATS company, what we have to do is add value beyond what they're doing. Utilize our AI data to take it one step further than what they're doing. So you, you guys even promote Winston, right? You promote, there's LinkedIn agents, there's agents across all different kinds of ATSs. But are they solving the problems that recruiters want them to solve? And are they manually eliminating the tasks that recruiters get bogged down in? And that's our goal is to ensure that those tasks become easy and an agent can fulfill them.
16:40 Speaker 01 Is there a specific industry that you guys like hone in on to ensure that you can really focus it industry and/or level? So you've got obviously high volume is much different than mid-level, et cetera, et cetera. Then you've got warehouse, you know, warehouse workers much different than, than healthcare workers. Do you guys hone in? How, how does that whole process work?
17:01 Speaker 04 Panda Logic, we use bots to qualify warehouse workers and delivery drivers. And all you had to do was ask. Three questions. If they answered them right, you send them a link to, you know, day one start and they show up at a place and they start work. That's not the same for high complex jobs, nursing jobs, right? Doctor jobs, um, trucking jobs even require certifications, right? So we're looking at jobs that have at least 500 applicants per job, which means there's a volume that we can accelerate and highly complex jobs. So it's high volume and, and highly complex jobs
17:42 Speaker 04 where we can extract all the data and the more data you extract, the better the qualification is for these higher level compliant types of jobs. So we're not talking about stand up jobs. We're talking about, you know, sit down jobs.
17:58 Speaker 00 Any idea why LinkedIn doesn't use LinkedIn to verify applications? I'm kidding. Let's move on to still move on to another question. So let's back to.
18:06 Speaker 01 Any ways we could talk about how LinkedIn doesn't use LinkedIn data to actually match shit? Cause they're, I mean, they're bad at all of it, to be quite frank.
18:15 Speaker 00 All right, Terry, back to your scary comment. Um, just something in the news this week that I just terrified the hell out of me. Elon obviously is in the news big time with firing federal employees. Write us an email. Let us know the five things you did this week. It came, it came out that he said he's going to use AI to throw all the, the content of the email in an AI and AI will decide whether you can keep your job or not. Is this the kind of future that employees can expect going forward that everything you do, everything you write, say, produce, et cetera, is going to get chunked into an AI? And the AI will recommend whether you keep your job or not. You're nodding your head yes, but I want you to expound upon that.
18:55 Speaker 04 Yeah. Well, my wife likes Elon and she asked me to name five things I did in the house last week and I.
19:03 Speaker 00 That's only because she owns four Cybertrucks, but that's a different, different podcast.
19:08 Speaker 04 But yes, I believe that now this one example of five things you did last week to validate and to either make a determination to fire somebody or keep somebody is kind of ridiculous. But I can tell you, my last performance reviews that we did all had chat GPT content in them. Right. We sent out questions. How did you write on these, this and that? And everybody went to chat GPT and I'm reading through performance reviews and they're all the same answers. Right. Because everybody took these performance questions, put them in chat GPT, took the content back. And that was their performance review. So how do you distinguish between anybody that's using a large language model?
19:54 Speaker 01 Outcomes.
19:55 Speaker 04 Right. So yeah, it comes down to what did they provide? What was the quality of data and some measure of timeframe? You can't just do one instance and make a hiring decision or a firing decision over that. It's got to be over time where this data is being collected and being evaluated. And I think performance reviews already reflect that in the world today. We're using AI to do performance reviews and we should.
20:24 Speaker 01 Well, we had an instance where GoFigure are one of our favorite companies on this show, Amazon. They used AI to advise and also to fire individuals who, you know, weren't, weren't performing. And many of these people were actually on vacation. So they weren't performing, of course, because they're on the goddamn beach because they were taking their time off. So I mean, it's, it's one of those things where we definitely need to, as you talk about humanizing AI. The, the shit that Joel's talking about is what everybody's thinking. They're like, oh my God, they're just going to throw this stuff in a blender and it's just going to go crazy versus being able to take a look at a job.
21:01 Speaker 01 A job, especially recruiting is a list of tasks and being able to see which of those tasks are administrative. And you can actually take off the back of a recruiter so they can do the things that they need to do. Like talking to candidates, talking to hiring managers, actually talking to people. So that the whole humanization piece to me really rings true.
21:23 Speaker 04 Yeah, here's the positive equivalent to your example with Elon. If you're doing upskill analysis of current employees and you have a new job description and you run that job description, you know, most employees are not updating their resumes and providing it to the company, but they're doing tasks each day. And if you look at those tasks, you compare it to a job description, we can analyze the skilling and the gap of that skilling and present learning modules to reduce that gap and enable internal employees to actually get higher quality jobs, higher paying jobs within the organization. And I think a lot of this is a positive aspect of using that same methodology to helping
22:08 Speaker 04 companies improve their hiring process by tapping into existing employees. And I think the world, you know, you guys know the labor market as well as I do, and there's, you know, underperformance of jobs. And I think if you can tap into first your existing employee base, identify a skill gap, get them trained up to that quality of job, you save yourself so much time, money and energy. You have somebody that already appreciates the culture, they want to continue working there. It becomes just a really positive approach to hiring candidates.
22:45 Speaker 01 Well, it seems interesting to me that we have had technology that can enrich resumes, right? For years now, where it just goes out on the web, looks at different points, enriches a profile and or resume. There's no reason why we can't do that for employees to be able to collect data on performance, right? To be able to enrich the actual performance profile of that individual so that you know and so that they know exactly what's happening day by day. So, I mean, I think there are some really good aspects. And as we take a look at Doge and the stupid shit that they're doing, if they could actually analyze performance outcomes, right, versus doing this stupid send me an email thing, at the end of the day, we could be more productive.
23:29 Speaker 01 And then obviously, the people that are, you know, they're shamming out and they're not doing as much work, they're not going to be able to hide as easy.
23:36 Speaker 04 Exactly.
23:37 Speaker 00 Terry, do you ever see a Rage Against the Machine moment with AI? I've mentioned on the show all the time that my 18 year old has more vinyl records than I do. There's a need by humans to want to go back in time or feel a certain way. AI is changing that how we apply to jobs and how we're upskilled is changing that. Any, any guesses from you or do we, at some point, do companies say, we have no AI whatsoever in how we hire? And that's, that's a certain attraction for a lot of candidates. Are we there yet or do you think we'll get there?
24:08 Speaker 04 No, but we got to get there. So one of the initiatives we're looking at, now we've been doing AI for 23 years. Generative AI is new in the marketplace and needs to be adopted. And a lot of the agentic agents are being generated from internal large language model usage. But I think that there's a problem with utilizing this, like think about deep fakes, right? And stealing somebody's picture, stealing somebody's voice and propagating that and creating a resume that's not authentic and it's not you. That's a problem. And that will be something that I think over time we'll be able to identify and dismiss. But it's a hard problem to solve.
24:53 Speaker 00 Chad and I are buying the Kinko's brand and we're bringing back resume printing and cover letter printing for the, for the kids. Faxing. For the kids, because they, they love, they love all that stuff. That's Terry Baker, boys and girls. Terry, thanks for joining us. For those that want to connect with you or learn more about Daxtra, where do you send them?
25:15 Speaker 04 Daxtra.com. It's pretty easy. D A X T R A dot com.
25:19 Speaker 00 Pretty easy. Thank you, Terry. Good to see you again, Chad. That is another one in the can. We out.
25:26 Speaker 05 Wow. Look at you. You made it through an entire episode of the Chad and Chase podcast. Or maybe you cheated and fast forwarded to the end. Either way, there's no doubt you wish you had that time back. Valuable time you could have used to buy a nutritious meal at Taco Bell. Enjoy a pour of your favorite whiskey. Or just watch big booty Latinas and bugfights on Tik Tok. No, you hung out with these two chuckleheads instead. Now go take a shower and wash off all the guilt. But save some soap because you'll be back. Like an awful train wreck, you can't look away.
26:09 Speaker 05 And like Chad's favorite western, you can't quit them either. We out.
Previous Rippling v Deel: Over the Top Next Shredded: Humly, Future Education, Appcast, Skillvue, Upwork, BrandLink, StackOne and More Page updated November 11, 2026 · transcript and summary are machine-generated