“The challenge isn’t just identifying skills, but ensuring the data is validated, diverse, and reflective of real-world performance.”
My guest for this episode is Greg Gasperin, CEO and Co-Founder of Merify, a skills data verification platform. Greg and I discuss the future of skills-based hiring, which we both agree requires a shift from traditional resumes and degree-based qualifications to more dynamic and validated assessments of a candidate’s Skills.
We discuss the fate of the skills based hiring movement as inexorably bound to the ability to have meaningful, quality evaluation of skills that is based on direct input from humans.
Greg discusses the genesis behind Merify based on the need for continuous feedback, data integrity, and community-based validation as essential components of modern talent evaluation.
Of course we also also cover the role of AI in automating skill taxonomy updates, ensuring that assessments remain relevant and aligned with industry trends.
Topics Covered:
- Skills-Based Hiring:
- Moving beyond traditional resumes and degree requirements to more accurate, validated skill assessments.
- The importance of diverse, peer-reviewed feedback in creating a trustworthy skills database.
- Community-Based Skill Validation:
- The role of continuous feedback and real-world performance data in building a dynamic skill assessment system.
- How internal talent management can foster trust in skills data before expanding to external hiring.
- AI in Skill Taxonomies:
- Leveraging AI to maintain up-to-date skill taxonomies and adapt to changing industry demands.
- Balancing efficiency with transparency and explainability in AI-driven decision-making.
Takeaways:
- Trust in Data: A reliable skills-based hiring system requires validated, peer-reviewed data that accurately reflects a candidate’s real-world abilities.
- Continuous Feedback: Regular, diverse feedback is essential for maintaining the accuracy and relevance of skill assessments.
- AI for Agility: AI can automate skill taxonomy updates, helping companies stay current with evolving industry needs.
- Start with Internal Trust: Focusing on internal talent management builds confidence in the system, paving the way for its broader application in external hiring.
- Transparency and Validation: A transparent, explainable system for skill validation is crucial to mitigating biases and fostering trust in AI-driven hiring decisions.
Articles Discussed in the “Take it or Leave it” Segment:
- “Leveraging Professional Education as a Bridge Between School and Career”
- Summary: This article from Fast Company explores the gap between academic learning and practical application in the workplace, emphasizing the need for higher education to incorporate more professional training to better prepare students for their careers.
- Discussion: Greg and Charles discuss the importance of exposure to various career paths early in education and the benefits of integrating professional skills training into higher education.
- Link: Fast Company Article
- “To Make the Most of Credentials, We Need a Better Marketplace”
- Summary: This article from Jobs for the Future highlights the need for a unified marketplace for credentials, where skills and credentials can be transparently evaluated and trusted by employers.
- Discussion: Greg and Charles talk about the importance of having a trusted, centralized system for credentialing and the role of employer buy-in for the success of a skills-based hiring approach.
- Link: Jobs for the Future Article
Full Transcript
PTW_Gasperin
Charles: Alright. Welcome to the show, Greg. How are you today?
Greg: Doing pretty good, Charles. Thanks for having me.
Charles: Yeah. Yeah. Thanks. Tell us a little bit about your background, what you’re doing, and your company.
Greg: Yeah. Thanks. Yeah. And I just wanna say I’ve been a fan on LinkedIn for a bit, so it’s been great to get to know you for the last couple of weeks, and thanks for having me
Charles: on. Mhmm.
Greg: So yeah. So I’ve been a software engineer for well, a little over twenty years now, so a long time. And yeah. And I never thought I would start a company, but here here I am. Took me a while to feel confident in my own skills.
I’m a pretty pretty anxious person, so I was definitely having difficulties around interviews and, you know, having to, you know, reprove my my skills from scratch at each each interview and then throwing that data away afterwards was not really an environment I could do well. Even if I was a good fit for a job, I would find myself getting really kind of panicked before interviews. And I was right. I was thinking it would be really nice if there was a way that I could rely on my track record. What I’ve proven in an actual work environment in the past, and that’s kind of the basis for verify.
There’s there’s not really a trusted way of capturing that data, and it goes beyond interviews where that can feel like its own skill that’s unrelated to job performance. It also is part of that part of the skills based movement where because of the same underlying lack of data, huge swaths of the qualified workforce are not able to participate. So that was the motivation for starting Yeah. Clarify.
Charles: Very cool. We’ll talk about that a little bit more, but I’m because that’s that’s a a lot of the substance of what I wanna talk about. But still, I feel like jeez, starting your own company is a pretty big deal. So but before that, So as a software engineer, did you have formalized training? Did you go to school?
Did you teach yourself? And what kind of engineering were you doing? Twenty years a lot has changed in twenty years in the field of software engine. Yeah.
Greg: And it’s a lot to keep up on for sure. The skills landscape is changing extremely rapidly, and it’s harder and harder to keep up on that. So I personally have gotten more and more kind of specialized in what I’ve been doing, so it’s less of a generalist and more of a specifically end to engineers so I can customize those technologies. Oh, yeah. I I have I do have all of the, you know, traditional markers for you know, I went went to got a traditional education and went the very, like, you know, standard path towards employment.
So yeah. But since then, I have definitely kept up on skills in my own time. Charles: Yeah. So did you study software engineering though in school? And
Greg: I did. Yeah. Yeah. I did. You know, I honestly just picked it because there wasn’t much exposure to what other opportunities that are out there.
And this actually, you know, we might talk about this in some of the articles that we’re talking about later, but there isn’t much exposure to what options there are for for employment. I was I knew about you know, I like the problem solving aspect of computer science of programming.
Yeah. And so I was like, well, I’m eighteen. I’ll just pick that as my entire career because that’s all the information I have.
So I went with it, and luckily, it worked out okay.
Charles: Yeah. So you didn’t have any prior experience as to how well, you might be able to do it or anything?
Greg: Nothing like that. Yeah. I I I definitely wish I had I had more of exposure to, like, what other options there are for you know, it’s just like, you know, I knew I, like, problem solving aspect of it, but there are a lot of jobs that do that involve that. And having, like, a, you know, maybe, like, a skill level understanding of what jobs require that would be Right. Would have been very useful information.
Charles: Yeah. Well, we’re starting to get a lot of those things now. Right? And even though with the just generative AI or something, you can find out a lot of career information on your own, but we sure didn’t have that stuff before. That’s for you know, that’s for sure.
I I had an interesting thing for me. I I always want to be an orthopedic surgeon when I was a kid. My dad was a psychologist. And so I went to I went to a pretty small, you know, liberal arts type school. Didn’t have a lot of undergrad.
Didn’t have a lot of, like, engineering or forestry or dairy site, you know, there’s all kinds of other things you can learn in school. So I went there to be premed and I I kinda sucked it well. I’m really good at science, but I found that I got a little bit screamish sometimes, and I’m like, hell, I think I’m a guy that’s really needs to be, you know, cutting into people quite honestly. And and at the same time, calculus was really hard for me. So I’m like,
Greg: I don’t know if this is not that.
Charles: And I just started taking psychology classes, and I did really well. So that there you there you have it.
Greg: Well, I think calculus and blood have driven a lot of people into psychiatry, I think.
Charles: Yes. A lot of calculus. What You know, yeah, I’m a psychologist, not a psychiatrist, but yes. Math runs the world, though. You know, calculus is sitting there behind everything that goes on.
Just the world’s a bunch of fractals, you know, and and sometimes those equations are running everything. It’s crazy when I start getting into the theory of math. But but I went to a a big school, a big university for grad school though, and it was really interesting because then I’m like, oh, they have forestry, they have kinesiology, they have all these other things. That I think I might have ended up potentially, you know, being really interested in. Remote sensing. I didn’t even know a remote sensing.
Greg: Mhmm. Yeah.
Charles: Do you know what it is? Remote I
Greg: don’t have no idea what I just But I don’t like the idea.
Charles: It’s satellites and stuff. Yeah. Being able to collect data from the earth from, you know, above the earth or whatever.
Greg: Okay.
Charles: Yeah. I thought, well, that might be really cool. Mechanical engineering, you know. So I didn’t have a lot of that exposure, but slightly different story, but but I still feel like I was a little narrow tracked maybe into into what ended up doing. Mhmm.
Now we have so many more tools available. Right? And and just even thinking about hiring. So so on your career, journey. You get out of college.
You start software engineering. Did you stay with the same company for a really long time? Did, you know
Greg: I did my first yeah. I stayed I started around the, you know, the dot comboom. So all of my friends went to dot com companies, and I went to a financial institution, which was a surprise even to me. But, you know, it’s I was able to stay there for for a bit. And it was an excellent development team.
But even just to to hit on the training that you get in higher education, our the company I I I was hired as they used our school as kind of like a starting ground for hiring people, but they recognized that a lot of the professional skills that they needed were not really something that what they were trained for. Right. So we our the first six months or maybe four months of that employment was all about being retrained for specific job skills for that role. And that would’ve it I feel like it’s it’s definitely a direction that a lot of higher education probably should be, you know, moving towards as specific job job skills training.
Charles: Well, the seed was sown for kind of what you’re doing now in some sense. Right? I mean, you had a really good opportunity to see that you don’t have to have a core set of skills that are required, you know, that you can you can if you have the the abilities and skills to learn things and to integrate into new things that you’re asked to do, then you don’t have to have the stuff maybe that the job actually requires fully. Right? I mean, that’s kinda what we’re operating now with a lot of a lot of the suppositions we’re making, you know.
So I’m just I’m moving. I’m slowly gonna get to we are slowly gonna get to talking about verify because I think it’s really cool, but we gotta get there progressively. So starting your company. Right? So what Look, it’s not an easy thing to do.
So how did you go from where you are working on a in a role to say, and you know what? I feel so strongly about this. I’m gonna put myself out there and then and how did you go about the process? One thing I’ll say, Well, that I know who’ve been involved in startups, spin founders, who have the ability to do the development and coding, have a massive advantage. Right? Because you don’t have to co pay somebody to do to do it all. I think that’s a huge help. So, you know, for me even stuff I’m working on, the technology side of it is always the barrier, the expanse, the time. I I lose control of the tech side of it, you know, fully. I guess when you’re
doing a lot of it yourself, that’s helpful.
So tell us a little bit about that transition, you know?
Greg: Yeah. There’s there’s some advantages and some disadvantages for sure. The advantage being it’s nice to be able to, you know, when you come up with the ideas and you get the customer feedback to be able to implement that pretty quickly. But you know, this is my first time out and the the more I delve into it, the more I’m understanding all the other roles that are required. So it’s I’ve heard a lot of, you know, the idea is an important the implementation is. And then, yeah, implementation is an important the the marketing and sales are. Right. And all of them are equally important. It’s not one versus the other. So if you have one, that’s great, but you need all of the components to to build something that successful.
Right. But it is nice to be able to implement ideas ends customer feedback as as it comes in.
Charles: Yeah. Prepared. Yeah. A hundred percent. I mean, I’ve I’ve been as a a part of a small force, most of my time on my own.
So I understand, you know, when I when I get people involved that have specific skills that help me be better, I mean, that’s what it’s all about. Right? And Yeah. And that’s what building’s all about. So you have a means now to look at people skills because you have your own platform.
Greg: So That’s right.
Charles: So look, we talk a lot about SkillBase hiring. Everybody talks a lot about it. I think there’s there we all agree. I don’t think there’s anybody who feels like, oh, this is a horrible idea. It makes a lot of intuitive sense.
It has multiple multiple wonderful things about it as far as access and, you know, about really having making sure that what’s required to do the job or to learn to do a job. It’s something that someone has not not the crap of a of a resume, etcetera. But it’s more than that. It’s not just let’s substitute a resume for an assessment. And I feel like that’s an easy way to think about it. But when you think about skill based hiring. I feel like we’ve also got, you know, the the kind of, hey, we have these tools that just look at a resume, look at a job description. They’re gonna tell you everything you need to know about how to match somebody There’s no substance there. There’s no real evaluation or verification. Right?
So Mhmm. To me, that’s the challenge. Yeah. We love skill based hiring. We can use AI to put a bunch of labels together and match those labels.
Is that really doing anything for us? I don’t know. What do you think?
Greg: Yeah. I I think that I think we’ve we’ve taken, like, the first step in the right direction, which is moving towards buy in. There is Yeah. Like you said, I think everyone is in agreement that this that skill space not only is a more fair practice, it allows everyone to participate. It’s done well, but it also provides more effective talent map mapping and more effective and streamlined works and more fair career paths, transparent career paths.
But that’s just one step of many. With I think I’m sure you’ve heard the really heavily reference Burning Glass Institute article that was talking about and and Harvard Business School that said that of the companies that dropped the degree requirement, only one in seven hundred roles was affected by that change in policies. So opportunity didn’t really change. And it’s it’s I think the most important thing to recognize here is that there there are a lot of components that you need to solve in tandem to make a reliable solution. There’s so many there’s so many underlying issues
that are at play here.
So, like, what’s the data source? If we’re not looking to more traditional markers for talent, like, degree is in title and tenure. What are we looking at? You know? Yeah.
How do we how do we access this full qualified workforce? What what was socially used? And then there’s the agility aspect of it. Like, how can that data source effectively keep up with the new like, new and quickly evolving skills. Do we need to evaluate each skill in a totally different way?
Or is there a more common unified way to do it that can naturally keep up with the pace skills. And then, of course, there’s the trust issue, which is, you know, around fitting. Like, how can we trust these results at all? Yeah. That’s obviously one of the bigger issues.
There’s badges, there’s certificates, there’s course work. Assessment testing, psychological assessments. And, of course, unfortunately, resumes, which are, as everyone knows, I think, the worst data in all of talent management, but they’re used to build reps. Sure. They’re mostly
Charles: Yeah. Yeah. I have so many stories. So there is a guy I’ve been doing this a long time. So there there was a guy probably twenty years ago.
And he reached out to me and he had another job. And if he was telling me he kinda lived in his office and he he was starting this company at night And his whole thing was what he called the integrated evidentiary resume. And it kinda was like a LinkedIn profile except you had other people rate you on there and you had assessments that would come in there. And and he wanted that to be a portable standard. This was way before LinkedIn.
And it was a genius idea. I mean, the guy was basically ten to fifteen years ahead of his time and didn’t have the you know, the clout to be able to move this thing. But the concept was amazing, and I helped him with some of the assessment stuff. And he, like, literally, was like, I live in my office. I do this job.
I do this all night, and then I work my regular job during the day. You made a lot of interesting people. You did this. I worked with a lot of startups and a lot of I a lot of people would, you know, ideas reach out to me, which I’ve always really liked. So, you know, people have been thinking about this in some way Mhmm.
For a while. And you know, it it’s a glacial thing though. If you just think about one person, one entity isn’t gonna just change this and make skill based hiring all already happen, you know, tomorrow or even next year. It’s gonna take a lot of different things coming together. Because ultimately, what we’re really talking about, in my mind, to make this work, is universally accepted standards protocols.
So what’s our skills taxonomy? How are we assessing this stuff? And it it seems insurmountable because I feel like everybody’s vying for that, you know, every company wants to have the assessment. Every company wants to be the one that provides the ontology or or the the taxonomy or whatever. And that it never happens that way.
But
Greg: Oh, yeah.
Charles: I was talking to Jason Taseko. I think I can’t remember how to pronounce his last name. And he’s the executive director of the t three initiative, you know, and we’ve talked about that. And I don’t know how much he’s told you about that or you’ve learned about if you’ve spoken with him or looked at the stuff that they put out, which is amazing. But they are working on and
a technology protocol and actually I forgot what he called it, but it’s essentially they wanna give a wallet to millions of learners and give them Mhmm.
Money to go virtual money to go off and buy training courses and add to their skill badging. So in other words and this is a layer that any company can tap into this. Right? So it’s agnostic to the company. It’s kind of like an infrastructure that any company can get involved in. I don’t know if you’ve heard about So that was the first big thing, and I probably did a very shitty job of explaining that. Is that something you’re familiar with?
Greg: Well, I mean, well, first, the the the the digital wallet concept is hugely important. Like, this is data that all candidates should own whether they build that data at a specific company Right. Or they build it externally, this is your career data and you should own it. So the there’s one important aspect that often gets overlooked that, like, I should be able to build my skill data at company a, and then use it to apply to company b, and then build it there and continue on my career. So that’s that’s that’s something that a a full solution needs to provide. But, yeah, the t three initiative. I’m kind of just getting familiar with it myself, but there is one project that they’re working on. And I forgot the exact name, but it’s essentially, like, stacked credentials that they’re looking at a community based validation source. And that’s kind of the the approach that we’ve taken. So trust obviously is the biggest one of the biggest issues in building something that’s adopted.
So what we’ve done at verify and just a little bit more on on on our focus. Our focus is actually currently, our focus is internal talent management, actually, right now. And the reason why we started there is because we want to build employer trust first. So if you do it internally and your process works identically for external candidates, so, like, talent acquisition and talent management work in the same way, then you have this implicit buy in from employers. That was one of the reasons.
And another reason is around how our metrics work to back this data. But, essentially, what we’ve done is that community based skill validation. So we start with a data source that is universally goal for every skill, which is first hand observation of skills being performed. Not only does that apply to any skill, whether it’s technical or non technical or any industry, but it also grows automatically with the changing skill landscape. So if we want something that has that agility baked into it, and then we take that data and back it up with metrics that that are automatically applied that are since we have this standardized data point, we can automatically apply these metrics that are around what is the weight of each data point? How important is each data point? And also, how do we have enough data? Do we have enough diversity of opinion? Enough confidence in the network of people providing feedback? Do we know enough about the reviewers to feel confident in the results?
That’s what we’re starting with, and we’re applying it internally. And then we’re also our long term vision is to also expand that out to more of a a full picture where you’re you have this scenario where you have this digital law that you take with you.
Charles: Yeah. So what you’re doing is I feel like a missing ingredient, which is some kind of verification. Right? If you’re just parsing apart a resume and saying, this this person has this skill because we interpreted that they’ve done these things. Sure.
You you can’t say there’s no way that’s Correct. But, boy, it sure would be nice to have some kind of money in the bank, some kind of way to to know that this was actually done. And at scale, right, I mean, the biggest issue with with hiring is just getting a signal, a quality signal, and
having the right thing to receive that signal. The rest of it’s all noise. Right? So the more that we can can actually have a a credible signal that’s parse apart into discrete things that represent something that we know and can believe they do. We can send that signal and then an employer can can have the ability to receive that signal. And now we have a lot more clarity and now we understand it. We we can really feel good about what we’re getting. I think internally it’s it’s gonna be similar.
For sure, I think more in hiring just because that’s been or my head’s been for so long, but but obviously skills or skills, you hire on them, you develop them, you you move through your career with them. Another piece I think is important for all this to really work is connecting the educational system. So if you start looking at ed tech, and I’m sure you’ve looked at a lot of ed tech stuff. And we might even be, you know, ancillary too or able to be involved in some of those conversations with people in that world about your product and stuff, but it should start there really. I mean, if we have the skilled language, imagine somebody, you know, coming out of the even high school or wherever it starts with some kind of again that wallet. Right? You you don’t have to give that wallet. I started getting an allowance when I was ten or so. You know, I had a wallet I didn’t use it very wisely and what, but, you know, I don’t I don’t have a couple bucks in there when I was ten. So I couldn’t have a and I was in the cub scouts and the weavers.
I got little badges there, you know. Yeah. I could still weave a basket. Somewhere, I guess.
Greg: Yeah. And I honestly, I think the Boy Scouts did did badges in the right way where you actually understood you need to understand the skills that make up those badges. I think they’re they kind of are, like, a skill level of badging system, which is something that you can very much get on board with. That’s essentially what the system is.
Charles: I was a boy scout for
Greg: a while.
Charles: I found in the cub scout that they let you just have your parents kind of do, you know, a lot of it for you. But then when I got to the boy scout’s, I did it myself, That was pretty I think
Greg: We actually originally were gonna call our company Meritage. Oh, really? That was that was kind of our our our model around the ties to the learning program. So if you supply and demand need to speak the same language, that’s at at its core. So if you have skill based workforce.
You understand the skills of each individual, and you have a skill based roles, definitions, or job applications, job postings. There then you are speaking the same language, and everything becomes much clear. It’s basically you’re operating at the most atomic level of saying like, okay. This is but this is how flexible we want our workforce to be, and now we know at the lowest level what we need and what people have. Yep.
Charles: And
Greg: what that allows you to do is to say, okay. I am I know I have skills a and b, and I know I want to get hired for this role that requires skill c or I want to get promoted internally for this role that requires skill c. It’s a very transparent process, and I know exactly what I need to do. I’m not
just taking a course for no reason. I’m taking a course specifically to get from where I am to where I wanna go.
Yeah. And then not only that, you can I mean, that would naturally translate to a course that provides that scaling and skill fee? If you have a if you have a consistent method of evaluating, so how I determine that I have skills a and b, but I can also use that same method to determine the effectiveness of upscaling programs. So say, for example, fifty percent of people who took this course improved x percent using the same underlying method, then you can have this really clear path of, like, okay, this is a course for me. This is this is how it is to get from a to b, and this is the course I need to take to do it.
It just makes the the whole process more transparent and open.
Charles: Right? So so and we talked about verification and you explained a little bit. Let’s let’s just, like, take me. I’m I’m in a company that uses your product. I have a I come in. I have some kind of so when I’m coming in, how do we know what skills I have to start with before anybody really rates me and then walk me through what data goes in to the equation to help verify that I have a particular skill, how all that works.
Greg: Yeah. Absolutely. So right now, we’re we’re working with companies that are rolling out things like a career progression frameworks. So we’re starting with we’re starting with companies. And once we once we get to the topic of AI, I can talk about a little bit more about skill text on vocabulary.
Who knows? Never heard of any of their I don’t know. So our focus is on the skill data validity. So so we’re we’re we’re a trust system essentially.
Charles: Right.
Greg: So what we do is we start with very simple in the flow of work feedback for that is given by your peers. So we want we essentially operate like a scientific study that’s kind of the core of our system. So Mhmm.
Charles: You
Greg: know, you would never trust a scientific study that had two data points. And from one person or for two people. No. So so we are saying that you need diversity of opinion to have a trusted data source. You need data points over time to have a trusted data source. And we’re making that continuous feedback very simple in our system. So you can provide feedback when a project ends or when your scheduled review time is or even as you notice behavior in the flow of work. And we conform to this very simple quantitative, almost like a check-in or a survey around the scale data, but we’re relying on this diversity of opinion, this as a starting point. But that’s just the starting point. The core of our system is what happens with that data.
So anytime we get a new data points, any new observation, we collect all the data for that candidate or all the data for that employee, every review they’ve ever gotten and all of the data we have about the people who provided that feedback, all of the data about the reviewers. So we’re not only trying to understand the candidate, but we’re trying to understand who provided the feedback and how important that is. So what we do with that is, first, we go through our interpretation metrics and those are designed to determine the impact of each individual data
point. So we consider all this data and then say ask very simple questions, like common sense questions, like how good is the data, the reviewer, and the skilled doing? How recently have you worked together?
We take a rolling standard deviation to determine the outlier. So does this data point represent an outlier? Did this person get three people to completely lie about their skills versus all the other data points we’re getting, which is saying something else? Does this reviewer give everyone an expert or everyone a beginner when they’re providing feedback? We take all of that to consideration to determine the impact of each data point.
And it’s a dynamic system. So say, for example, today you reviewed me, and the system doesn’t know anything about you. Very little impact. But a few months from now, you can actually fill out your verify profile, and we’re fine we find out you’re really good at one of the skills you’re reviewing. We’ll retroactively adjust the impact that you’ve had on it because we now know more about all the people involved.
And then there’s also this whole series. I end up I end up talking too much about this. I end up going forward of our system. I’m a date data nerd, but we also have our vetting metrics, which are used to determine how confident we are in the results. So naturally, there’s this confidence rating to say, like, okay.
Do we have enough understanding to feel confident in these specific results. So do we have enough, for example, data overall to vet you at your current scale level? Is there enough diversity of feedback in your network for us to feel confident in the overall results? And my favorite one, number one is to do we have a do enough people agree with your current skill level. And this one really speaks to me because it allows us to essentially combat impostor syndrome. People who are not necessarily self promoters who may, maybe, you know, underrepresented groups traditionally have more impostor syndrome. And what we do is basically say, you think you’re performing at an intermediate level in these skills your community and our interpretation of that data says you’re actually performing at advanced or even expert will automatically move you up to that level and and notify you of that fact because you should feel confident in your abilities in that. And then, of course, the opposite is true if someone took a five minute course and they think they’re an expert, we’ll move them down to the more appropriate level.
Charles: Yeah. That’s awesome. So you have a lot of quality control from a lot of different nickels. Because the first thing you think about is Oh, well, like, when you ask somebody to give you a reference, you know. No.
No. You’re not gonna ask people who aren’t gonna give you a good reference, you know. Like, my cup scout crocheting, Merritt badges, mostly my mom crocheting that thing. Right? Yeah. When I turn it in, I get the little thing. So is it as meaningful? I can’t really crochet very well. But, you know, when you have all those checks and balances, I feel like that’s that to me is what you have to be able to believe in these things. And a multi greater sources, you know, I feel like a really strong I think a lot of the other and there’s many different approaches, but when I think about credentialing and you mostly think about you’re gonna take a test, you’re gonna get, you know, a certain score in that test, you’re gonna get a little metal that says, you know this subject area and you can carry that around with you, which is not necessarily a bad thing at all. And I would imagine that tests and assessments would be a great supplemental Mhmm. Data source for for you all to have, you know.
Greg: Absolutely. It should be a very, like, kind of, choose your your your own adventure kind of situation, both from the candidate or employee side and from the employer side. So if we have this modular system that says, okay, as a candidate, I want my profile to be verify’s core data about me. And I also want to include these tests that also conform to the skill space interface or whatever interface I’m building. And then I also have these psychological assessments that also conform to it.
I’d piece together this full profile that is a full picture of me as candidate or me as an employee. I can even include things a lot of companies I met along the way in the past couple of years are doing preferences at work. That’s something that could also be used to conform to the same interface, and I’ve built my own profile. And the same is true from the employer side. I can say, like, okay.
These are the data sources that matter to me. And I only wanna see Greg’s profile with the verify core data and this assessment test, not this one. And I’ve picked and if you have your data set up in a flexible enough way where you can slice and dice it in real time, then you can say, like, okay, as an employer, I care about this. And now I I I’m looking at the data that matters to me.
Charles: From a real you know, where my mind just went is, oh, well, you know, we we’ve been doing this for years with tests in in one way or another, but say you have an assessment of soft skills, let’s say. Right? And then you have somebody who’s ready to buy all these other people on their soft skills. The assessment is an objective proven, you know, reliable measure of those soft skills that’d be really interesting to see, like, okay, there should be a really high correlation here. Right?
Like, it should give you some additional understanding, but as a test guy, I also wanna know that my test correlates with these other ratings of similar or very different things and doesn’t correlate with different things. That’s what you really want. Right? That’s it. That’s really interesting because we’ve been doing that
Greg: It’s all additive. Yeah.
Charles: Yeah. Yeah. We’re doing it a long time. Like, when you’re validating a test, you make sure it measures what it’s supposed to, you’ll you’ll take you know, hopefully, if you can get the the the same people to take three different tests that are similar, plus years, three is just an arbitrary number, and two tests that are highly different. So if it’s personality, also take some cognitive tests.
And then you have this thing, a a multi trait, multi method matrix where you look at all the stuff and you’re like, oh, yes, my test is supposed to measure this. It correlated very highly with other tests that are supposed to measure this. Didn’t correlate with tests that don’t measure this. So you get a lot of of confidence that your test is measuring what it’s supposed to, you know.
Greg: Yeah. It sounds it sounds very much like the checks and balances in in our system. It’s like very much like you have to get multiple data points to feel confident in in in the overall results.
Charles: Yeah. That makes sense. Like you said, you’re gonna trust a you’re gonna trust a paper that has, you know, two data points from one person or one person. Right? Not not as a
Greg: Yeah.
Charles: Not as a as
Greg: a Which is
Charles: Yeah.
Greg: Which is how, you know, most most internal talent management works. It’s, like, my opinion and my manager’s opinion and then Yes. And it happens once a year. So, like, how can we trust those results?
Charles: Yeah. We all hate those. It’s, like, everybody hates performance reviews. Everybody hates resumes. Talk about getting rid of them.
There’s been some really good stuff that people have done to get rid of them, but they’re such structural things. I’ll tell another story from my recent thing. Probably about six years ago or so, so a lot has changed, but I had the chance to work with somebody who was doing a grant funded program called the essential competencies program. And what it was was essentially getting we tried to get about ten or fifteen employers. It wasn’t we we ended up with, like, five. Five big global employers to say, yes, we are willing to choose a job, and we are willing to forgo the resume and all this other crap, and we’re just gonna give these candidates an assessment, and we’re gonna hire the candidates just based on their assessment scores. And then down the line, we’re gonna look at how those people did relative to the people do in a regular process. What an amazing idea? It was awesome for me to be part of it. And what happened? What do you think happened?
Greg: I don’t wanna hazard a guess because I I I sometimes people are good at assessment tests and they’re not good at applying those skills. But if the assessment test is good at is good at detecting, you know, applied skills and it could have the opposite result.
Charles: Yeah. Well, I wish that we could have gotten there. What happened was not a single one of the companies was actually operationally able to to give into the idea that they had change their hiring process for one job. The structural way they had it all set up, even though we had leaders engaged, it was just it all fell apart whenever we had to say, no. No. No. You’re gonna substitute these assessments for in your ATS, you know, it was just it was impossible. So while the intentions were there, it was very difficult to make that happen. So I think you start to see how it’s hard in an organization to even support this because we’ve got this all this HR tech infrastructure that is really hard to change it all at once. You know? I think your HR tech stack, you’re hiring tech stack as different pieces have been added on it, different times, a lot of times, and it’s not always just easy to make a change across the board, you know. So Mhmm. That’s I think I’ve got to deal with.
Greg: There’s the you know, on the topic of adoption, I think it’s a similar scenario if we can, you know, take those assessment tests and, you know, sometimes it’s hard for to get an existing employee to take an assessment test, but I’ve done that in the past for sure. And also, you know, any method of scale evaluation. And we apply that internally and employers are seeing that pay that. I know that this person is good at this business and this and this test confirmed that or this other process confirmed that, then you can start to get some buy in. And even if you start using that for your internal talent management process, like, even better, if you prove it internally on
data, you know the odds of I feel the odds of them applying that externally increase because they got this built in confidence.
And we’ve seen that with our with our first customer, they are interested in we we used our process internally, and we started talking about, you know, using it for the hiring process as well.
Charles: Oh, yeah. Very cool. So did you I’m you might think I’m changing the topic, but Exactly. Are you in Olympics? Scott, did you watch any Olympics?
Greg: I think the only thing I saw was the Australian Brake dancer. Oh, yes. Yeah. That was that.
Charles: Yeah. I watched a lot of Olympics. I’ve always liked it, but the streaming capabilities that peacock had where you could just surf around and watch any event at any time. Basically, you know, that had already happened or live in some cases. I watched a lot of Olympics. A lot of stuff. There’s a lot of things I didn’t even know existed. And it was amazing. I thought I was really well done and it was the first time and that, you know, a lot of the different tech shows and podcasts I listen to, people kind of said the same thing that the ability to completely be untethered from this is what event you have to watch at what time, but just being able to use the, you know, the remote and just move through the the app, the streaming app and watch whatever was great. But I did kinda have the the idea, well, the Olympics are kinda like skills based hiring in some sense.
Right? You can be from any background. As long as you’re able to, like, prove it on the you know, in whatever tests of skill that you have. Right? You can make the team and then you get trained on the skills and where your skills are weak.
You get more training. Right? And you get focused training. And then you get to compete, and it doesn’t matter if you’re from, you know, Zambia or New York City. There were even examples of people.
I think there was one American cyclist. I can’t remember her name. She really didn’t have you know, some of these people trained since they’re like, too. Right? She really didn’t even have a a a long history of it.
She just took up she was a VC person or he is, and she took up, like, cycling for fitness and turned out she was really, really good. And she kept winning races. So she got bound her way on the US team and then nobody even she won a medal. Nobody even had her a medal. Nobody even not really who she was.
She just came from out of nowhere. So, you know, you got the skills and you’re allowed to compete. And then you get a medal. And that medal is a personal or portable badge. I mean, if you’re if you’re an Olympic in general, it’s something you can be proud of your whole life. If you want a medal, you know, you get credibility anywhere you go. People recognize it universally, and it means something. So I don’t know. I was like, yeah.
Greg: That’s a that’s a good analogy because yeah. Totally. It’s an excellent analogy because it’s that is if you’re if you’re using markers for getting people to the Olympics or getting people into a job that aren’t necessary necessarily getting capturing the entire available audience than you’re missing out on a lot of talent. Yeah. And the challenge is around, you know, identifying where those skills are.
Charles: Yeah. Yeah. I mean, it
Greg: it and and
Charles: I could I could shoot holes in what I just said too because I’ve been trained to be very skeptical. But but in general, I thought it was, you know, it was good and maybe I just spent too much time watching that stuff. But I kind of felt a little bit of a sense of loss because for two weeks, I come home, from work and, you know, figure out what sport happened today? What do I wanna watch? And powerlifting, you know, skateboarding, surfing?
It was so much good stuff. So I’ll I’ll You might
Greg: have to go back and check some of that.
Charles: You should. Man, totally should. I I there’s there’s a lot. So good. Well, let’s see how about we hadn’t talked much about AI.
So one thing I’ll I’ll relay from that about AI in the Olympics. And then then Maybe we’ll do the take it or leave it, and we’ll come back and and kinda have our AI say. But the Right. One thing I will say while we’re still on the Olympics. There was at the very beginning, there was a an ad. Google had a bunch of ads. That’s the one crappy sign about it is I had to watch a lot of ads. Every time you fast forward it in like a three hour event passed a little orange node, you had they they penalized you with a hundred and twenty seconds of ads, you know, like I’ve already tried to fast forward through the content and skipped over some of the ads. They accrued those and made you go through them no matter what. But Google had a lot of ads.
The very, very early one because I think people got pissed off and they took it off and it was It was very interesting and telling it was I can’t remember her full name. It was American Sprinter Sydney. I can’t remember her last name, but the whole commercial was this little girl is running track and she’s idolizes this woman and wants to write her a fan letter. And so the dad’s like, okay. We’ll just use AI to do that.
And, like, you know, the the Google Gemini wrote the letter and then she, like, copied it, you know? And and so everybody’s, like, Wait. That’s like a special thing a kid’s supposed to do. Yeah. AI to help kids write these emotional letters, you know.
So it was just a really interesting and Google never even thought about it that way. Right? I think a lot of other people saw it that way, but
Greg: Who else? You know,
Charles: keep in that, you know?
Greg: Maybe that Kim was writing a letter to, like, fifty Olympians and, you know, she just needed some productivity increase. Yeah. I’m so Well,
Charles: and it’s a lot yeah. It’s a lot harder for her to start with a blank sheet of paper. Right? But But I don’t think that they showed him her, you know, erasing some of it and then, like, changing it around, you know. Anyway, the concept is, you know, that’s a real danger of what we have is that, you know, people are just gonna lose a lot of these Like, you’re gonna lose on experiences, etcetera.
But at the same time, she wrote a damn good letter, probably better than with seven year olds. Right? So It’s a
Greg: du alice Yeah. Suspiciously good. Yeah.
Charles: We yeah. These are du alleys of AI. So So let’s take a look at these articles here.
Greg: Oh, yeah. Can we do the Fast Company one first? Because I think that really leads into the other one pretty well.
Charles: Yeah. Yeah. For sure. So so the first article is from Fast Company was a couple months ago, leveraging professional education as a bridge between school and the career. Right? So it’s talking a little bit here about that gap and how are people closing that gap and, you know, is it actually worthwhile doing? So so talk, what did you think
Greg: Yeah.
Charles: About this article?
Greg: Well, I very much agree with the core proposal of the article. Definitely often thought higher education needs to do more to prepare students to apply specific skills on the job. And that that has been my experience too. I, you know, started that my first job with this training, of course, where I where I learned a lot more applicable skills in that short period than I might have at a few years prior at school. So, you know, if that so it kind of it kind of brings I don’t know. It kind of begs a couple of questions, which may be a hot take or not. But if the goal of higher ed is to prepare students for the workforce, then why isn’t professional education kind of the majority of their focus? You know, I think it would be beneficial for students and employers if the academic side of hiring it was focused on exposure to different careers, maybe early on, something that we talked about before. But then there should be a real focus on if that is the main goal of higher ed to prepare you for the workforce. There should definitely be this component of professional education that’s baked in.
The other question though that comes out of that, which is maybe more controversial as why aren’t companies subsidizing higher education? This article talked about developing course work with a specific employer and which is great. But, you know, if it’s you know, if if this is I’m I’m the one benefiting from a a skilled workforce, you know, subsidate subsidizing that benefit that I’m getting feels like something that makes a lot of sense. And if you’re working for a community college, probably, like, right now, like, hey, we already do both of those. And, yeah, absolutely. That’s I personally think that the community college model is kind of how higher education should progress.
Charles: Howard Bauchner: Yeah. It’s really interesting. I think just overall, it starts to we have to start questioning what is higher education about and who is it for in the form that we think about it. So I brought the story earlier. I mean, I went to a small liberal arts college. I didn’t even know a lot of this stuff existed. Right? And everybody that I went to school with pretty much became a banker, a doctor, a lawyer, you know, something like that. There wasn’t a lot of diversity of careers, to quite honestly.
Greg: Mhmm.
Charles: I might be the most diverse out of at least the people I know. But But at the end of the day, I’d like I would love to be a real estate agent, honestly, in some sense, like, I’m really good at it. I’ve I’ve renovated houses in bottoms, sold them, and I really enjoy that whole process. And I feel like if I had known about that more and how that worked when I was younger. That might be a feel I’m actually would have gone into.
Right? So just the exposure and access, I think, is really important. And again, thirty, forty years ago, it’s a different story, a lot to change. But in terms of just general awareness. But I agree. And I think some some you can get that. Right? But you’re in, say, a community college. Well, then are you worried about the prestige of, hey, I went to a community college and not a general four year college? Oh, I don’t know.
That might people might look down on me for that. So there’s I think there’s a lot of of this. I I feel like educational institutions are diversifying. I feel like, you know, and there
Greg: is more
Charles: you can do. And they have online classes and stuff, but but I feel like the connection outside of the core things that we’ve all been doing in college and universities for a long time would be great to have. So
Greg: Yeah. Absolutely. And I think there’s there’s also a a corollary to the skills based conversation that we we’ve been having. You know, you brought up, maybe you would have been interested in being a real estate agent or may and I brought up earlier, maybe I would have been interested in some other field where problem solving was kind of the main skill. If we understand, again, here or even in education, if we understand interests, say, at the skill level, then we can more accurately map students to what career paths would make sense for them and give give them exposure to those career paths.
I think that would be I would have left that to to have been an option when I was in school.
Charles: Yeah. Yeah. For sure. So I think we agree on that. I’m gonna give it a thumbs up. I’m gonna take that one.
Greg: I think a thumbs up. Yeah. Definitely. I I also love the articles that Fast Company has been coming out with on this topic, especially recently. There was a really great one about neuroscience behind effective feedback.
No one’s on that. I recommend that one.
Charles: I did not see that one. I would like to be right one for them I’ve been studying on. How do I get that how do I get that done in my spare time. I think you could submit ideas and stuff, you know. Alright.
So Yeah. I think so. I was looking at I used chat TGP to say, like, what media outlet based on what I wanna talk about would be best for me or whatever. So I think it was wired and fast happening or two of them. Yeah.
And they have they’ll tell you all the different stuff you gotta do. It’s it’s in my to do folder. Greg: I only use JBT to write to Olympias, so I don’t
Charles: know. Exactly. I use it for a lot of stuff. But the the article I’m working on right now is one that I have earmarked to maybe, like, kinda ratchet up and see if I can get it through, but you never know. Anyway so our second one here, and we’ll let you go too.
This one is from jobs for the future, which is and it’s from twenty twenty three, Halloween twenty twenty three, but jobs for the future, I feel like is pretty pretty, you know, doing some some good stuff here. So this article is call to make the most of credentials, we need a better marketplace. So we talked about that a little bit already. So what’s your take on this one? This one basically?
Yeah. Summarize what you think it said and then give us your take.
Greg: Oh, man. This covered so much. I think it might be a little difficult summer guys. You know, they really crammed a lot into into a couple pages. But I think the the bridge from the last article that I to this one that I was really interested is the the last article mentioned you know, people might look to someone with a real estate license to find certain skills like, you know, customer service, communication, time management.
And I think that plays in well with this article because we’re talking about credentials. And one of the in my opinion, coming into this article, one of the most important things about credentials is we need to understand what skills make up that credential. So for example, I don’t think many people looking for customer service or to fill a customer successful are specifically looking to people with real estate licenses. You would look at a lower level at the skills that make up that license. So I think part a really credential a really important part of credentials is understanding the skills that make up that credential.
And the another really important part is employer buy and invest with everything else. Charles: Yeah.
Greg: If employers don’t trust a credential, then there’s no incentive to earn it. So those are my two thoughts going into this article, and I think the article did a really great job of covering both of those.
Charles: Yeah. Yeah. Absolutely. So it it does cover a lot of the stuff we’ve talked about, and I think it uncovers some of the some of the issues. Right?
Is that anybody can throw a credential around. Even if you verify it, if if it doesn’t meet a standard or is accepted, it makes sense, and we know where it came from, etcetera, it’s not gonna be as it’s not gonna be acceptable. So we have to have this kind of unified marketplace, unified venue, where people can exchange this information, it’s absolutely critical. And just think about what we talked about earlier, so much different than, okay, we’re gonna you’re gonna use an assessment instead of your resume to help you get hired. Yeah.
That’s part of it. That’s great. But that’s such a small minded thing. And and even we’re having trouble getting people to do that. It’s an amazing an amazing thing if we can get that executed. But in the grand scheme of things, it’s such a small dinky part of what we truly think about as skill space hiring. And this is some of the, you know, stuff we’ve got a sleigh to do it. But The
idea of these, you know, learning and and employment records, LERs, and some of the stuff the t three initiative is doing with creating the technology layer to bind all this stuff together. Right? Is a is a huge factor, as I said.
I I think without some kind of neutral party that is got everybody’s best interest in mind that can
kinda be a foundation of this. It’ll be hard to make it work just with private Mhmm. Companies. Right? And it’s Yeah.
Everybody’s gotta be on board.
Greg: Totally agree. It’s earlier. And I think I think one way we can support that, like, understanding is I I personally think that in order for a credentialing marketplace to work. There needs to be a trusted underlying skills marketplace. That’s, like, oh, these these are this is the data that makes those credentials.
And we we can we can undetermine our buy in based on how trusted that data that core data is. Yeah. But, yeah, absolutely. I think something like this, they’re what they’re proposing makes a whole lot of sense. Yeah.
And there needs to be a way to consolidate all of the data sources that are out there in a way that stakeholders can come to, like we talked about before, stakeholders can come to and say, these are the data source that I care about, and this is why.
Charles: Yeah. Makes total sense. Good deal. Wow. Okay.
So let’s see. Let’s get back here. Do we even have time to talk about AI? We’re almost out of time. So how let’s look at it.
Let’s do it this way, because we could do four shows just on AI. We’ve mentioned earlier, you know, hey, maybe we’ll have some AI in our platform. How do you envision? And I guess, are you AI free now besides machine learning? Algorithms.
And and, you know, tell us a little bit about that and what you might think for the future. What’s the use case for AI and what you’re doing?
Greg: Yeah. Absolutely. So AI brings up, you know, we mentioned before trust in data and agility and data source, and there’s also, you know, bias considerations. Transparency is also another concern around skills based practices, and AI brings up a couple issues there. I’m really glad that there’s now a shift towards transparent AI because, you know, probably driven by some of the losses related to bias in AI.
But there if you can’t explain why you’re making candidate suggestions or internal promotion suggestions, then you leave yourself open to to issues like that. And so the reason why we started with an algorithmic solution is because we were our algorithm is our metrics are based on really common sense, things that we actually understand. Like, how good is the reviewer in a skill, how recently we work together, things like that that I that’s that’s not nuanced. It’s very clear. And also, it’ll it naturally builds this explainable, transparent data system.
So we can we don’t have to do any extra work to say, this is why we refer a candidate or this is why we we think this person is a good fit for promotion. So there is this move in a transparent AI to to resolve that, but it is a hurdle for AI. The thing that I really am excited to apply to our system apply AI to our system is around skill techs on me. For me, and this is our our near term product of ours. One great AI application is to automate the maintenance of skilled taxonomies. So skilled taxonomies are extremely complex and they’re nuanced, which is a great problem space for AI. And maintaining that manually is a ton of work. Especially given the pace of evolving skills. So we have this concept of related skills in verify, but I did a simple test with chat JPT to ask what skills in our systems are related to, for example, sales. And the result were spot on, and they even came with explanations for why those relationships were.
And it didn’t take any maintenance on my part. It’s based on the data available global data available to it. So I think that’s an excellent application for AI. In our system, specifically.
Charles: Very cool. Yeah. I think there’s applications for AI in a lot of places, but Greg: Oh, sure.
Charles: Feel like I’m glad that you didn’t build your house on AI. From scratch. Right? Like, there’s a lot of other stuff that can go into it before you start really thinking about it. And I think taxonomic structures and, you know, being able to make those relational relational associations are are really important.
The difference being, again, that you have a way to verify it, it’s not you’re just not taking the the the taxonomy maps onto what you’re already doing, so you you can connect those two with some evidence. Right? So good. Yeah. Well, thanks so much for for your time today and for Sharon has been great talking to you, and appreciate your explain what you’re doing. And I’ve heard all my silly questions. So No.
Greg: Those are great. Thank you very much for for having me on. It’s a real pleasure.