Where Artificial Intelligence and Machine Learning will really make a difference in staffing
What areas can AI impact at my staffing company?
Tune in to find out why AI and machine learning have become indispensable in many areas.
Hear Beeline CEO Doug Leeby, among others, talk about the importance of AI in staffing.
Learn how AI and machine learning can be leveraged to build a bigger & faster talent pipeline.
Sign up to check out the full session for in-depth insights that you can apply to your own company!
Jan Jedlinski: Hey everyone we are back. Thank you so much for joining in on the panel. We have Jason, Marcus and Doug here. Who will be talking about artificial intelligence and where it makes difference for staffing. So I'll be handing over to Jason now, enjoy the session. If you have questions, ask them on the chat and guys enjoy the session.
Jason Ezratty: Thank you, Jan. And thank you, not only for inviting me and us to this session, but for having the entirety of the summit, it's been a great thing to look forward to and to hear from colleagues around is one of the things that these things do, it forces everyone to reach out to each other and say hi. Before we com ents m with introducing the topic and asking some questions of my panelists, I want to make sure that you're aware of exactly how esteemed these panelists are.
Gentlemen I'd like if you don't mind give a quick introduction of yourself so everyone can fully appreciate who they're dealing with. Marcus. Can you kick us off?
Marcus Sawyerr: Yeah, sure. Happy to do so, thanks Jason. And always good to see you as well Doug. Just to give you a bit of background, I've been in the staffing and recruiting industry last 15, 16 years spent a lot of time with careerbuilder.com where I rose through the ranks.
So you could say from, inside sales all the way through to run the staffing and recruitment division across Europe, and then moved to the Adecco group about five years ago. I was responsible for digital transformation to start as employee number one, and then really focused on buying and building invest, been in HR technology companies for the group.
we spent a fair amount in two years. we acquired. nine companies and we invested in different organizations as well and scouted about 1,048 HR tech companies. And I was president of DECA group X at that point. Long journey in a staffing and recruiting industry and just embarked on a new journey.
Actually as of today, you could say founder and CEO of EQ community, that's focused on bringing in multicultural professionals into the world of technology and empowering companies to become more diverse and more inclusive. So really happy to be here.
Jason Ezratty: Thank you, Marcus. and you're one of the rare breed individuals.
I, I think like Doug as well, who, who can bring all of the perspectives having an understanding how we build it, how we sell it. How we plan for it, how it becomes a reality. and I couldn't be more excited about your latest endeavor. so best of luck with all of that and ask Doug for anything he can do to help. Doug you can bring for your intro.
Doug Leeby: I got a lot of slack on LinkedIn for confessing, my insecurity of being on a panel that is esteem, but I don't feel nearly as esteemed. And I'm hoping Jason can join us. He's going to try to dial in. So potentially he'll get on and brings a wealth of information, especially from the staffing side. But just quick intro December 5th, mark, my 20th year with Beeline. and I think people have a pretty good understanding. Most of you probably hate us and if we usually extract, so I won't go into anything about being steamed, but two decades in and looking forward to being in this panel with you folks,
Jason Ezratty: Doug yeah, Jason Heilman coming in from a Bullhorn would be with us shortly.
He's our, the headliner. So we wanted to bring him out in that last letter.
My name is Jason Ezratty. I serve as a chief data scientist for Brightfield. I'm also a co-founder. And my background started in the world of of data science. We can call it data science in the world of biology and trying to predict things like when a drug might cause a heart attack or when certain things might imply a pollution event.
And then I found my way in staffing it around the year 2000 have been a part of this fun journey of And then increasingly take advantage of data.
In this topic. it's my favorite and least favorite topic of all. It's my favorite topic, because it's my job. It's what I do every day. It's what I think about leveraging data to find patterns. but it's my least favorite because I think it's broadly misunderstood. We always say artificial intelligence as an area of science or technology. is really describing how something gets done. It's not the what itself. and so I find it's better to talk about the, like, why are we talking about these things and then how can AI solve that problem?
Jason Heilman: Hello. Sorry. I think that happened did not want to Jason's on this call. Bad. Got it. Hopefully you can hear me alright.
Jason Ezratty: We've already prepped the audience to know that you were just the main headliner and the most important to bring out. So we want it to let, like the opening act.
Jason Heilman: Alright , I'm going to take my turn to prep the audience to say that's completely untrue.
I am just here right along the ride, because I didn't talk to you guys.
Jason Ezratty: Yeah. And I guess to break, to bring you into the fray, I'll I'll bring my boring intro to ask you to give a little background about yourself.
Jason Heilman: Yeah, absolutely. So I am Jason Heilman. I'm the SVP of product automation, AI with Bullhorn. I come to Bullhorn through the acquisition of Herefish, which was a tool that is primarily focused on automation.
Technology focused on the staffing burden. But we do quite a bit in general. We're in a back end where we're going to subtly power, a lot of what we do through AI and prior to that I was with a number of other or sorry, not a number, another applicant tracking system.
So I've been in the industry since 2006.
Jason Ezratty: Thank you. And thanks for joining us, I think is a great selection of people to bring interesting perspectives to, to the same types of topics. So in, in my intro, other than bemoaning the nature of the topic that we're talking about how in, in light of a why the other thing that's important is when we talk about artificial intelligence, we're really talking about two things at once.
One is artificial. We call it artificial intelligence because we expect technology to living behavior. Usually we mean human behavior. So the idea that on the other side of technology, whether it's be a text via, some other form of chat. That we could believe that's a person there. But I think the 80% of what we call AI in this world is not trying to mimic a human intelligence on the other side of an interface where it's really just trying to mimic human intelligence of how the brain works. So understanding, we say like machine learning understanding data patterns and how can tell us what's more likely to happen in one case versus another case.
So we'll probably be talking a bit fluidly between those and we'll do our best to be clear about which type we mean. Cause they don't mean the same thing. to jump into the topic at hand I guess we'll start Jason, if you don't mind what is your first memory of using the term artificial intelligence?
You go as far back as you can.
Jason Heilman: I, if I had to think about it, I don't know the exact day, but I think its gotta be in sci fi I think it's gotta be a movie. Maybe it was Terminator. That's probably my initial introduction into the concept of AI. If we move it over into business world I think, again, I don't know the exact timing, but it feels like it was shortly after the craze of big data.
We went in, everything was about data. That was the headline for two years maybe. And then it seemed one day it's snapped and all of a sudden everything was all about AI. And maybe that was the evolution of learning with big data. That's the only way to process it, but I don't have an individual moment, but that feels like where it came from for me.
Jason Ezratty: Marcus, Doug, do either of you have a sort of an aha moment or a spark memory of when you started. Either you thinking about these terms as a kid or more professionally when it became part of the job and not just something in a, in an external headline kind of thing.
Doug Leeby: At the risk of sounding arrogant, I coined the phrase many years ago, long time ago. So again, I want to let these guys have their due. And Marcus, maybe you remember when I taught that to you many years ago, but I don't know what the date was.
Jason Ezratty: Marcus?
Marcus Sawyerr: Yeah. I think it was I was born in 1984, so I think around that time, was when I got a message from Doug and my parents had mentioned to me that there's this guy called Mr. Leeby, and he's coined this term. And if you follow this path one day, you may be on an AI and I'm still proud of.
Yeah. So I followed that journey and got there, but I think one of the other times that I've heard about AI. And it was probably in about 2008, actually, when we started career builders to really start talking about recommendation engines. And there was a question whether or not recommendation engine was AI or not, and what that actually meant.
And a lot of it was really a similar look up and we see it in a lot of consumer products and services that we use, but in a job board world, that's really, when I started to come into AI and understand at least what it was. And I think the term recommendation engine is more it's more freely used in AI now, but at the time it wasn't called AI.
It was just a recommendation engine. But I think that was probably one of the first times I started to pick that up.
Jason Ezratty: When we think about even just the last 10 years when you thinking about like the era of big data, like you're saying, and there's been, a lot of different marketing terms and cycles of. Jason, I come back to you, specially given your record, which, which are the areas that you think were really the hype lived up to, or the reality lived up to the hype and so applications of what we're calling AI in either definition, the purpose of the promise and then obviously we'll ask about those that perhaps didn't go ahead.
Jason Heilman: Yeah, absolutely. For me, I think, and again, I think this is probably a lot to do with kind of some of the paths that we chose. But I think the automation, which often is conflated in with AI is an area where it is living up to that hype but it's not, it's maybe not as exciting and as sexy as AI, but I think what it does a good job of being able to allow people to easily configure business logic in a way that they don't have to have a computer science degree.
So to me, that's the area where, and again, probably because of where I'm at, where I've seen the most success, because oftentimes. I've seen that people are a little bit scared of what AI will do cause they don't actually know what it's going to do. It's you're leaving a decision open to something that you haven't fully trust with.
So I'm sure we'll talk about how trust will evolve over time. And many of the things that look to be very promising in the industry. But for me, that's why I feel like staffing firms in particular just really want to have control over the experiences that they're creating and by giving tools that make it easy to create those that's just where we've seen a lot of success and we see that kind of time and time again.
Jason Ezratty: Sure. And so if we look let's tunnel into that and get a bit more recruitment focused and so Doug, from your perspective, is there an area of the recruitment life cycle that you think is just completely inappropriate to outsource to AI? That must be part of the human experience.
Doug Leeby: I do. And I have to challenge my thinking because I sometimes wonder is this just old-school thinking and perhaps I can't break the paradigm and I'm wrong, but I don't know, but I will tell you I, I think there's a lot of promise in the life cycle, even in terms of matching, but I was thinking of it this way.
And you said earlier, a lot of marketing is out there, like we're AI powered and all that stuff. So I would imagine you were single looking for a partner and you went to one of these dating sites and it had this proprietary algorithm is going to match you. Hands down and be like, yeah, no problem. Let's roll.
Let's move like it, this doesn't work for me. So I think it'd be great from a vetting perspective to start to narrow the list. But personally, I still want to look somebody in the eye and I really want to go into the cultural aspects and the softer skills. And I believe there's certain components of that can be enhanced or at least informed.
But personally, I still struggle with this notion that it would just replace all human element. I just, it's not that my head is more, I don't think we'll ever get there.
Jason Ezratty: Thank you and thank you for this move, the powerful, take the glasses off to make. yeah. Thank you, Doug. That was excellent. Same question, but in the opposite unless you want to just agree with Doug but which are the aspects of I'll leave there for disagreement. Go right ahead.
Marcus Sawyerr: Yeah, so to be fair there are parts of Doug's statement, unfortunately, that I do agree with. I feel like there are pieces that when it comes to the human elements and the human aspects which are critical and crucial to making that match. And if you think about where AI could outperform humans, and I think the AI can outperform humans in many areas, it's really going to be where there's quantitative information.
So numbers, they are everything associated around that. So get into a shortlist. I really think that AI will be able to outperform humans now getting to that final list and pick it and understanding that cultural fit between organizations and doing that final match. I really feel that's where the humans will continue the humans.
And I'm one as well. We'll continue to add value over a long period of time. One other thing that I think is important because it's not only the recruitment aspect, but the sales aspect. So I think that it's really hard to automate an enterprise sales deal. And I think that will be hard to automate for a long period of time because they have different needs.
And I also think it's a missed opportunity. If you do try to meet that because staying close to your customer and really understanding the problem of working back is something will help you to continue to build out those algorithms. So I think if it's high frequency, high volume, highly repetitive, then as well, automation.
If there's a human element and aspect into it, my cultural match that's needed. I think that will continue to remain for some period of time.
Jason Ezratty: I think the extension question there and really curious for, and for any of you to jump in for this, is it sounds like we're assuming that the AI or the automation has to be complete.
Certainly if you have someone that's behind the controls in either of the situations you're talking about, I would imagine that there's elements of AI and leveraging data and probability to help that person with that decision, even if it's not making the final decision. I always say like one of the best examples of the success of AI is spam filtering.
Can you imagine how much more spammy email we'd be getting? Even not from the recruiters out there. Sorry. I don't know any extensions they're building AI to augment the human, as opposed to replace the human.
Jason Heilman: Yeah. I I'll jump in. I couldn't agree more. We, sometimes we even frame it as, it's your virtual assistant, right? So it is what helped you to be more successful and more productive. And I think as we look at where the hope I'm not jumping ahead here, but as I think about where the industry might go over tonight, I personally believe, and I think I, I think Doug agrees in Marcus you do too. The human is always going to be a very important element in the process. This is one of the most personal decisions that somebody could make. One of the most important, like what they do every day. Doug's been doing it for 20 years. And if somebody's an AI might not have been able to pick that job.
And it's such an important decision for someone's life. So I think the humans are always going to be involved, but I definitely think the role of a recruiter in an organization is going to start to change a little bit. They're going to be much more focused on maybe being more of a career advisor, handling those interpersonal relationships and being there to support someone as they're making these decisions or as they have questions that come along.
So I definitely think there's going to be a shift in the way that we interact with candidates powered by AI automation, whatever the given technology is, that's really going to make it better for everybody.
Jason Ezratty: And better inlcuding not just less expensive and faster but better because it's more nuanced and more finesse than what, unfortunately, the current world of recruitment often offers how often do we have to read stats on how quickly the human is scanning resume and really playing a numbers game.
Jason Heilman: So that to me is the main conflict between our automation versus human. It's when the humans trying to act like a machine, that's when you know it deserves to be automated. It's when the human is bringing more of their EQ to Marcus's journey. Then I think it makes sense. Speaking about some of the current reality for any of us in the data business, I think everyone on this panel as we roll through the, from the start to what is, hopefully now, hopefully the mid, the middle toward end of what we're living through in this global pandemic.
Has this been the greatest thing or the worst thing? For those in the numbers, prediction and forecasting business how have you guys seen this? Has this been something that makes models get thrown away or something that makes models more valuable?
Doug Leeby: By this, are you saying, sorry,
Jason Ezratty: COVID-19 global. The massive disruption to a world that was operating one way. So we've been collecting data and making probabilities and assumptions and training models on that data. And then all of a sudden, within a three-week period, at least outside of China, the world changed. I had to deal with that. I'm curious to hear how you guys dealt with it.
Or skip the question.
Marcus Sawyerr: Yeah. yeah. And maybe just to add a little bit on that, and just generally from a business standpoint, I think that whatever playbook you had around March, April that just got torn up. and you have to start again. And I think that one of the it's obviously a challenge, but for everybody involved, but everybody was involved.
So there was an understanding that I think within days you're going to continuously get anomalies misses going back to the human aspect. We can take that into context and understand the situation and where we are around that. That was programmed by an AI. And you were trying to make forcasts and there was no human aspect that could have been out control, whatever a business application that you were really spending time on or getting your information from. So I think that this is just another point where it's important to have the human connected with the AI or the predictions or the data qualitative and quantitative information.
As a combination is more powerful than one in isolation.
Jason Ezratty: I agree, I think was one of the more important things that we learned through it was we built AI to, to discover context. So certainly, if AI can think about context across thousands of variables at once, then shouldn't that be the best way to do it.
So I think that was some of the positives. The problem is if you didn't build your AI in, in light of global pandemic and what was driving it, then you're missing key assumptions and inputs to how that model should be thinking. And so that, that's where the human intuition I think comes in to test those assumptions and make sure that you can flex in the moment.
And anyone else want to talk about how this has impacted your understanding of data and how it can tell the future?
Jason Heilman: Maybe I could jump in just a little bit, I can't say specifically about thoughts I've got on the data side, but what I think has been great is the business side. So like every other business in a world at adoption has been accelerated these tools in the staffing industry.
Yeah, I think because I think we all believe in what these tools offer and the promise that we're making. That is a good thing for the industry as a whole. And ultimately, as long as all of the purveyors of these types of technologies continue to invest in. With this newfound interest, continue to invest in building up their products.
We continue to be able to be configurable for whatever the future may bring. These few things I think, align together. So we're all going to grow. I think ultimately it's helping the staffing industry to mature more quickly into these types of tools and as long as the tools and best back and ensuring with maybe more frenetic world, that's changing more quickly. All those kind of hopefully work together to create a better future for everybody.
Doug Leeby: I think that's a key point, Jason, on the adoption, when I think of it in that context Jason's right. I get this question a lot about what does COVID run in the context of COVID itself? I'm not sure that's created a lot of change, but economic downturn often spawns a lot of activities, especially around efficiency.
I think Jason in line with what you're talking about adoption, I see far more receptivity from our clients to look at what might be scary in the past, in terms of AI or machine learning, driven in the hopes of driving greater efficiency.
Jason Heilman: Yep.
Jason Ezratty: We touched on this a bit before, but my experience is when you talk about AI, you get two immediate reactions. Excitement and fear. Excitement we know what to do with this. Why most of us at this point are using it in some form of marketing.
But I think we often just sidestep the fear. Why do you think fear is still a part of it and are we not addressing it the right way? Doug.
Doug Leeby: I think it's a major part of it. I think it stems from an awareness, lack of education. You've mentioned some of the sci fi films Elon Musk, preaching doom and all the scary things that we hear the Chinese are doing. They're so far ahead of the negative connotations associated with these medications are terrifying. I think Jason, you talk a lot in terms of it just being about predictability and data and things that maybe we were doing in the past, just in a far more elegant, accelerated fashion.
So I think the excitement comes from wow, what can be done, but also, oh my gosh, did I just lose my job? And everybody's talking about human replacement and I think most of us, if not all in this whole form, not just on the screen, on the stage today, understand it's not about that. So that's my take on it.
I think the more we talk about it and we'll show practical answers to the, so what questions and you still have your job because it's cognitive intelligence versus replacement intelligence. I think we'll see higher receptivity.
Jason Ezratty: That rings true to me. it's disruptive by definition. Disruption has changed.
Change is scary. So I think that makes sense. And we could probably all do a better job of what's on the other side of that change. Do you, one of you guys, Marcus or Jason, have something to add there. Go ahead.
Marcus Sawyerr: Yeah, I was just going to say to Doug's point it's really bringing people along the journey, right?
If you're not bringing whether it's your staff internally or it's new hires on the gender of your onboarding them, we've got start onboarding people into AI and figuring out how they can leverage. And so a lot of the fear comes from, if you're a Turkey, you're not going to vote for Christmas, are you right?
So if you've caught up, you're going to get eaten and your lunch. you're you're you're you're going to be on the top of blood. You're not gonna, you're not gonna want to drive it forward, now if you understand the, okay, I'm the one that's also going to be empowered by leveraging the AI. And this is why I think we talk about human superpowers.
How do you connect people to an a AI or to that real quantitative information to give them their superpower. And I think thinking in that context versus the doom and gloom a lot of fear mongering will really help to bring people along with Jen, but people don't like, and they hear AI sounds scary because they're not in it.
But once you get in it and you start learning, you feel like you've got this real adopted. And I think all of us, because it's every single day when it comes to AI. Education is a key to it and make sure you're being part of that journey.
Jason Ezratty: Thank you, Marcus. And while we have you Think about, the important topic of bias especially in light of what you are founding.
As a data scientist, I hear that word bias and I think about probabilities and, biases, what we do for a living. But we also know that positive and negative, unconscious bias in hiring decisions. And so here we talking about, that kind of bias. So I guess to ask it in a different way than we often hear in the newspaper, do you have ideas or thoughts on where, AI and machine learning is being used to effectively combat unconscious bias?
Or do we really only at this point have more of the examples of where it creates a negative reality.
Marcus Sawyerr: I'll take that one, if you don't mind. Yeah. Yeah. Yeah. Yeah. Yeah. I think there's two sides to this. I was actually speaking to a company earlier this week. One of the things that they've done it they've implemented an assessment. And as part of that assessment, what they're finding is that there are people from diverse backgrounds, multicultural, black people, different natures that are really now starting to go through the process a lot further, get to interview and actually get hired.
And wow, this is great. We're getting more people, being the CVs and the time that supply it at this point. And it's you take this test and then you go through now, the percentage of people that they are now hiring is still low in correlate in relation it's relative. So in relation to the population so there was a much bigger problem, meaning they just the hiring manager, weren't looking at the people because of their names.
Now you solve. Excellent. And your top quality you're going to get in. So that's great. The challenge that you have is really over a long period of time. How do you evolve the algorithms to ensure that everything's taken into context to the population taken into context, the inherent bias is started to be removed. So I think that there's a lot of work to be done when building the AI.
So I see there's some positive pieces, but there's still a lot of work. I don't think we have the governance models around it because it's so new. And again, if you're not on the forefront and you're not on the front foot and being proactive about this, you're just going to be somebody that is going to be put in a situation where you're gonna have to deal with the algorithms cause they're already there and they're already built. So I think there needs to be more governance and more structure around the the AI that's being built on to who's basically responsible for it.
Jason Ezratty: And if I can hold you just for one other point on that, who do you, where do you think that governance should come from?
Marcus Sawyerr: Yeah. so I think it's collective. I think that there should be a governance not only at the company level, but there should be with oversight of those companies as well. I was, yeah, I was talking to somebody over there. We were just talking about the internet and the fact that you can access this information instead of having to read an encyclopedia anymore.
And we forget that. And then you forget the time when you didn't have to wear a seat. In your car, but we look back on that and we think, oh, that's crazy. Or people are on airplane smoking. That's absolutely crazy. And I think AI is AI similar to that when you're building algorithms and there's no governance we're going to, we're going to be thinking 20.
That was crazy. How could you do that? So I think that those governance models as they come at a company level, but you're going to need some governance bodies around it, but making sure that they are not slowing down innovation, because that's the challenge, the balance between driving innovation and making sure that you've got a structure around it. Yeah, company and boards as well, but I don't have the, I don't have the, I wouldn't say that I have got the answer now.
Jason Heilman: Yeah,
Jason Ezratty: Go ahead, Jason.
Jason Heilman: I feel like that's right. Yeah, I think right now we're in the world where we're in an airplane got our seatbelt on smoking cigarettes right now. And going from zero to one is where we used to be frankly. We'd looked at opportunities where we can add value in this arena, but it's been difficult. There's some ontologies that you can look to for keywords that can identify. But a lot of the baseline data in some ways, intentionally doesn't include some of the information you would need in order to build up a model for success.
I think it's there, but it really does take, people like Marcus and others, putting some focus in getting first step where others can build on and can continue to iterate and make that action effective. Stop smoking on airplanes.
Marcus Sawyerr: That's a good idea
Jason Ezratty: Marcus you have too many one-liners to try to vote, but I'm just going to stick with Turkey on christmas. Doug I know this is a topic you're passionate about. I think this is one that's worth going all the way around. And anything we're missing or opportunities or other perspectives you bring.
Doug Leeby: I just think we talked earlier, oops, I don't want you to throw away thing, but yes,
it's a difficult topic to, to discern and figure out.
And when I think about AI, we talked earlier about it's predicated on data. And so part of the problem we have right now and the murder of Mr. Floyd over the summer, it created this massive worldwide consciousness that we've got to do something. It's disgusting. It's taken this long to reach a tipping point, and this is just rates.
We have gender, we have sexual activity, all these different areas we have to look at, but the reality is we don't really have the data on it. And then I think there's a lot of fear of could I quit that VMS? Could I put that in an applicant tracking system and will I get sued? And how do I look at that?
Because if we can get comfortable and promote self-identification as such in these areas and people trust us. Then we can start to utilize what we're talking about, machine learning, etc, to extract and learn and present opportunities. But right now it's just a blank canvas. So it's difficult.
So when folks talk to us about, we really want to increase diversity, maybe on the contingent side, I think you have a baseline, so that's great. We need to go after that, but what's the baseline and they're pulling away from a data perspective to do that. So I think just socially, there's a lot of work we have to do with each other to get to that point where we can get to start getting to data points.
Jason Ezratty: Yeah, completely. Especially since I think it was some of the points that Marcus was making, that it doesn't have to be as plain as I see this person, I identify them as blank and I hold these biases consciously or unconsciously. It's also all the other things that, that are hidden, that may be variables in our system that, that we create models.
I think we often say the first step is acknowledging that these biases exist because that's how these models work. So the whole point look for bias, whether or not they're the biases we want to hone in on is the important question.
Okay we had bit of a funky time start, so I want to be conscious of just, not rolling too far but one of the things I wanted to get from you guys it is, yeah. If we suspend disbelief and forgetting about the history of how we got here, everyone's bored about history stuff, whatever suspend disbelief, and look forward five short years. Which do you guys think are the pieces that will look most radically different tomorrow versus today as a result of AI and if possible, not just automation, but AI is possible with any cases.
Jason Heilman: I can jump in. I referred to it a little bit, but I think even within five years, the role of the recruiter is going to look very different. I think that position will be massively important to an organization, but it will be important for different reasons that it's important today. But more specifically, I, I think prescreening applicant management kind of database managers. How you interact with your audiences? I think those are the things that are gonna look very different. The things, if you know the marketing, if you look at the top of the funnel of those interactions that are very time consuming and extremely inefficient today, and often inefficient in a way that isn't about losing time but losing quality and not finding the right people.
So that to me seems like the area where we got just massive amounts of opportunity to improve the system in five years are really reasonable timeframe.
Jason Ezratty: And just one quick follow-up with you, Jason there. So if the, if this title recruiter will remain, but the job will be different. Is it the same profile of person or is it a different profile of person?
Jason Heilman: If I go all the way to what I want to see, yeah, I think it is a different profile. So today I think the people that we look for are people with a sales background that are going to be very high energy, engaging, trying to persuade is persuasion is a very important role in that process. I think it'll look more I don't want, what's the right, just some of them maybe more nurturing, caring, I think not the recruiters today aren't that, but I think it will be less of a sales role and more of like customer success where you're proactively working and being an advocate. Maybe what it's think of it for your kind of your candidates and long term.
Jason Ezratty: Yeah, no, thank you. And I completely agree. And I think that's going to, I think that's one of the gate limiting the factors of how this works is. You're expecting different people to be behind the keyboard. With different skills.
Doug Leeby: Yeah, I think vantage points a bit more from the buy side. I think we were fond of saying the resumes dead and the job descriptions are dead and it's a little bit hyperbole today, but I think we're absolutely going in that direction. And so one of the fundamental problems we have is that these job descriptions, they're just not. And that's why MSPs are so important.
Why they're sitting there by selecting all the attributes are necessary and having these calls with the recruiters and it's just not efficient. So to the extent that we will deploy AI to do a better job, extracting what those attributes are, what success looks like for that outcome versus abiding by generic overall directional job description.
I think there'll be tremendous efficiencies for everybody in that. And then there's already good work happening there. So I think within five years we'll see a completely different approach to job descriptions. And I don't think you'll be using resumes the ways we do either.
Jason Ezratty: Well said you're here. Marcus, tell us how the world will most likely be most different.
Marcus Sawyerr: I really think fundamentally it'd be different with matching. And the reason I think so is that there's been an emergence of platforms. And when you have the emergence of platforms, not only do you have the input, but. You have the outcome as well.
So being able to close that feedback loop over time to figure out who performed well in a position and who performed well in a job will allow you to feed the front end of that. Somebody told me a long time ago that a job of a recruiter was to really put around holes or around pets, should I say in square holes. And that's what you do as a recruiter. And I think you're gonna see less of that. And I think you're going to see more fits, right? So you're going to see more round pegs, more round holes, and you're going to see that to be refined more because of the closing of the feedback loop and understanding what works. So I think the data was key and we didn't really touch so much on that, but garbage in, garbage out. And if you understand the outcomes and how people are performing, then you're going to be able to train the models better. And I just see that infinitely continuing to get better.
Jason Ezratty: Perfect. But I completely agree. And I think actually what you're saying in what Jason from Bullhorn is saying go hand in hand. The less you, the less your peg is round and more square. The more you have to sell someone that it is other skillset finding more roundedness for the round and that it makes sense for the inherent benefit of it. So better fulfilling the mission. I think everyone wins from that candidate, buyer, everyone. That's a great point. So we have a few minutes left. I've been a hog. I'd be curious if you guys have any questions for each other. I'm happy to open it up.
Doug Leeby: I don't know about you. I'm looking at the notes here that say get off the stage. I'm going to respect Jan, cause I think he's done an extraordinary job putting this together and I'm going to be quiet, but Jason, if you want to creep poetically, waxing, that's up to you my friend.
Jason Ezratty: I was using the time we were asked what is. No, you don't have time for that for sure. I will just round us out and thank I'll go back to where I started to this esteemed colleague Marcus. Thank you for the the Turkey doesn't vote for chicken. I'll be Jason very quick. I appreciate all of your comments and we look forward to everything coming out of your laboratory. And Doug I appreciate everything that you do.
Marcus Sawyerr: Are you hot in that? It was on my mind before.
Doug Leeby: Let me just explain something, Mr. Millenial. This is a sports code to the audience.
Jason Heilman: Just so you guys
Jason Ezratty: are aware. I'm pretty sure we're still on. Okay.
Thank you Jan.
Marcus Sawyerr: We appreciate it.
Jason Ezratty: Take care.