How Fama Lens Brings Policy-Based AI to Social Media Screening

[00:00:10] Hey, this is William Tidcup, and you are listening to the You Should Know podcast. We've been on today. Ben and I have known each other long enough to, well, to talk about our families and other things like that. Yeah, things like that. Things like that. Hello. Hello. I'm back. I'm back. I so love your mustache. I can't grow one, so I have to, you know. What are you talking about, dude? You've got one literally right there with some beard around. Yeah, but that's as thick as it gets. It just gets longer and uglier.

[00:00:40] So the density of it. You've got density, Ben. Surprisingly, though, a lot of maintenance. A lot of maintenance. I had to buy one of those trimmers and do it every time I shave. So it's, every time I do it, I'm like, I'm about to cut half of this thing off. I'm just kind of, oh, how did it happen, honey? I don't know. There's no class. There's no class. And you remember when you're younger, like you have your dad teaching you how to shave. Right. This way, go against the grain. Like, don't do this part. No one taught me this time. I had to learn. No.

[00:01:10] On my own. No, I taught both my sons and they both kind of laughed at me while I was teaching them. So I'm like, okay. All right. So you know what? YouTube it. Just do what you do to your generation. It's okay. AI, dude. AI could do it for you. You know, they were for a period. This is probably 15, 16. I had a long beard. I remember like a ZZ top. Okay. So what they don't tell you about long beards. Uh, very difficult to eat in.

[00:01:40] Very difficult soup, chili. Yep. See the bitches are certain things. It's just like, yeah, this is, uh, this is not good. Uh, yeah, I found myself washing my beard more than washing my hair. It's like, did you, did you condition it? Oh yeah. Oh yeah. I had, I had beard. There was a beard gel or beard cream that kind of keeps it glistening if you will, but just healthy.

[00:02:06] Uh, sometimes, uh, sometimes if I'm really thirsty and I drink a lot of water at all at once, like I, you know, you take a big cup of water. It's like, I'll have enough water in my stash that I can be like, and I can like drink a little bit of, I hope we're on video right now for this podcast. So just in case you didn't catch on, it's what I go. Yes. And I kind of like suck it down from my mustache. Drink the water.

[00:02:33] Got another, got another sip in another quarter, quarter ounce. It just kind of stored it there for me until I was ready. Yeah. Yeah. Yeah. He is somewhere in there, uh, suck, sucking down the water from a stash. Yeah. See, these are the things that we learned on podcast that we would learn elsewhere. Are we on podcast right now? Oh yeah. Well, we are, but Ryan will probably cut it. That's what he does. Thanks for that. And it's, um, all right, do me a favor and introduce yourself and your company.

[00:03:00] Thank you. Uh, I am Ben Monas. I'm the founder and CEO of a company called Fama. We, uh, created the category of online screening about 11 years ago or so. We bring the dimension of who people are online into the talent screening process. Call us like the Gen Z background check, uh, social media background check, uh, talent candidate screening, bunch of words that'll SEO well, but the end of the day, we're here to help you identify the behaviors you've cared long about.

[00:03:30] But in 2026, they appear online. That's what happens. That's what we do. And so sometimes, sometimes people are the same offline than as they are online. And sometimes they're different people. Sometimes they're different. Yeah. I remember you giving me an example of, uh, somebody's, somebody social that were, they were just flaming every, every, every, almost every company that ever interacted with Verizon, Chili's,

[00:03:57] it's just like just flame, flame, flame. And they're Apollo, they're applying for a job like in customer service. It's like, okay. So like, dude, okay. One is not true. Either you should never be in customer service or, or this is just your bit. Like, this is like, if it's a comedic bit, okay, that's one thing. But if you're, I'm, I'm tired of this type of service. It's not a, it's okay. All right.

[00:04:23] Yeah. Yeah. Yeah. Yeah. Or like the principal that said like racist things about their students, uh, the doctor that was selling body parts on Twitter. I think I told you about that one. Yeah. Crazy stories. Like I got, I got a whole bunch of them and at HR tech, we can talk about a mean person. People will post, you know, people will post crazy stuff and it's like, okay, is that like, is that true? Or is it, is that a bit like I'm not, cause I like, I like comedy or

[00:04:51] standup comedy in particular improv, uh, even more than that. I can't, sometimes I can't discern. Like, is that, is that the, is it a bit? Like if it's a bit, okay. Like I kind of liked a bit, but if that's not a bit, I watch the news these days and I'm like, is this a bit or is this, I can't watch, I can't watch the news. I can't, I got like a 15 minute. It doesn't matter what news Bloomberg, I mean, like whatever news pick a channel doesn't make a difference. I got 15

[00:05:20] minutes max. That's it. Yeah. Yeah. Yeah. It's a, it seems like it's like a bit half the time from what I'm going to show. What what's new at follow? Tell me, uh, give us the, the, the peak into, uh, what you're doing in the last quarter, but also what's kind of coming into 27. Yeah. Yeah. A lot, a lot is new, a lot that we're super excited about. I'll, I'll, uh, you know, kind of take you back, uh, you know, when we started the company to sort of frame what's

[00:05:45] coming, but you know, when I started Fama 2015 and, and truly it was like me, I think I even still have the pieces of paper where we were like drawing the initial UI, but like, you know, it was very much when you're selling a new product in a new category and introducing a new type of work, like you don't have an existing competitor feature matrix that you want to be better than, right. You can't just be like, Oh, like we're going to do what, you know, workday is doing and try to do it better and faster and using it. It wasn't

[00:06:13] like that at all. Right. And so a lot of the initial crafting, traditional screening products wouldn't that, that doesn't, it's, I mean, it's relevant to know them, of course, but it's, they don't do the same thing. There were parallels, right. But like background checks, the guy, I used to say were like, uh, you know, the absence of information, right. It was the value, right. You would say, Oh, well, if someone doesn't have a hit, it was all very hit based. Is there something there? Is there a hit on a felony? Is there a, you

[00:06:41] know, hit on a verification that's incorrect or something? And so I always say a lot of it was us trying to figure out like, what did customers want to do? How'd they want to run the product? How did they want to interact with it? And so we took like a, was very popular at the time, seemingly less so now, but a very, you know, design thinking approach where we're like, all right, what does the product do? When did you want to use it? Where would this button be? Oh yeah. You would want to have a feature that does that. Oh, you wanted this configuration, this customization. And so a lot of it was just like working with prospects

[00:07:09] of people who wanted to use it, who'd had a problem of maybe like missing something critical on a new hire that was obvious online. Um, and sort of how they were running it manually. A lot of people did this, like Googling people, Facebook, right. People still look folks up on LinkedIn. Recruiters, recruiters would do this, like this, this bit where I'm not really looking cause I don't want it to bias me. However, I don't want to burn my hiring manager with somebody that's toxic. So it's like, I'd always laugh at recruiters cause they're

[00:07:38] like, I don't, I'm not really, but I am. And of course they're all looking, they would all back then, they were all looking people up. Yeah. EEO says you like can't unring the bell. You can't unsee that sort of protecting class information, just like you said. So yeah, there were these kinds of like classic opportunities for software to come in and disrupt. Think like, uh, people doing things manually, writing it in a spreadsheet, emailing that in Outlook, doing a writeup in Word. You know, it was like

[00:08:05] perfect opportunity to generate like a cloud-based single point of access software solutions. That's what we did. Right. And, and it was super successful. We introduced this, uh, concept of like flags and keywords, uh, flags and keywords, meaning that we like built a set of AI base flags that would read like text image and video, just like a person can think like for my tech folks out there, like a set of classifiers that we built that would allow

[00:08:30] us to like identify intolerance, identify harassment and tolerance, you know, a threat, someone saying difference between like, uh, you know, my boss and I are going to kill it on a project this weekend versus like when I'm done with this project this weekend, I'm gonna kill my boss, you know, two very different statements. Right. Customers want to know about the ladder. Okay. There's a context layer that's on top of this, right? And that was what the technology was, right? It was like, and, and deterministic AI, we like

[00:08:56] hired PhDs in statistics who were good at things like semantic batching and all sorts of other types of very nuanced text image and video based analyses. And then, you know, we, we built these flags and we edited the flags when cannabis was legalized around the country. We split like cannabis from our other drugs flag. Right. And so that then allowed people to turn cannabis off or turn, you know, uh, harassment off if they didn't want to screen for that for whatever reason. Right. And so we also introduced this idea of like, and this is little product speak,

[00:09:26] but we introduced this idea of like keywords, right? So you could see, like, I want to know how people are talking about my company. If they're, you know, bad mouthing my leadership team. Uh, if there's a specific, you know, industry specific risk you want to identify, I'll give you one example. Like we were hiring, uh, uh, with a big UK, uh, government agency in 2020, we were helping, uh, vet the nurses who were giving the job who were, you know, in 2020, probably COVID was the thing that happened by the way. I don't know if anyone remembers, but like six years ago, there

[00:09:55] was this big pandemic and all this stuff happened. We all started working differently. Uh, at the time, you know, we were helping, uh, this big hospital organization, like that, the, the, the nurses that were giving the job, they wanted to look at like, uh, people who were anti-vaxxers or who had these like crazy conspiracy. I can't, I shouldn't even say conspiracy theory in 2020. No, no, it's just flat around these crazy ideas about, about COVID and what it was. Uh, and so they would use keywords. Like I want to see how people are talking about Bill Gates and WHO, because there

[00:10:23] was a whole conspiracy theory around, you know, Bill Gates creating COVID and a WHO conspiracy and all that. Uh, but he didn't, but you know, anyway, uh, that's what they wanted to look at. So it was a perfect use case for keywords. And so, you know, over the years, the way that Fama kind of emerged and it like created an industry and competitors and people that did the same thing and UIs that looked very similar to ours, which was very, very flattering, but you know, uh, users would kind of

[00:10:50] sit between like their own policy that they had, which was essentially like, uh, you know, something that they had created that like, yeah, you know, there's this many, you know, sorts of intolerance, these keywords, these things, we're going to flag it. Otherwise we're going to clear it. And so just like every other software product, customers would take our flags and keywords, their own policy, smush those two things together. And they would kind of sit in the middle as this

[00:11:15] like translation layer. Right. And what that meant was like a lot of time. If you have a lot of keywords and a lot of flags reviewing stuff and going through information, right. And trying to like discern, even in a world where we were taking, you know, 99% of posts out of play, but you still have for some people, a lot of content to review when you're out there. Right. And there were also like false positives that kind of, as you said, like could generate legal exposure, you know,

[00:11:44] if a customer may be like, like, Oh, I don't want to know about that. And so in any event, um, AI is changing the game as we all know. It's, it's like crazy. I mean, I'm just, COVID was real. AI is real. It's just, okay. I just want to make sure that we're on it. COVID was real. AI is, is real. Like I was a daughter too. Trust me. Like I've been in the game my whole career and I was like, this is another like web three crypto. Everyone's

[00:12:10] going to have a coin type metaverse. Yeah. Metaverse. Right. And I was like, it's the same, same, you know what it turns out. It's not, it turns out it's like a real thing. And it's actually like crazy, I think, uh, new services economy coming. That's my hot take, but, uh, side, side, that's a side note. Uh, general idea here is like, we started looking at the art of the possible over the course of the past year or so. And we're actually introducing

[00:12:37] a new paradigm in social media screening. This is our big announcement. This is like one of the first times I'm talking about this. Oh my God. So I'm jazzed up. Our team's excited. So in the same way that we introduced flags and keywords, 2015, we're now introducing something called policy-based screening. Policy-based screening is a concept. Our first product we bring to market

[00:13:02] is called Fama Lend. And what that allows customers to do is to tape their governance policy. And a lot of people don't even have a governance policy. If they do have one, it's like intentionally vague, intentionally general to protect the company. Right. That's how most policies are structured in the world of HR as you know, very well. Um, so a lot of it is like, okay, we create a policy with

[00:13:27] a customer and then we will actually screen against that policy. Well, essentially deconstruct a policy in plain language that a customer creates, uh, convert that into effectively what is like a multi-agent consensus architecture, which basically means that agents do stuff based on a policy and they all have to agree with each other before they end up, uh, you know, making a decision to add or

[00:13:52] remove something from a report. So we're still not scoring people. We're not giving you like a yes or a no, it's not a thumbs up or a thumbs down, but instead what this is, is like a fully auditable, fully observable, fully traceable AI system. And look in the world of like AI and hiring, I think all of us, I'm not even going to name names or go into the various ongoing lawsuits that are happening right now, but it's a very tricky area to get right. And so short story is we

[00:14:19] basically put the controls in place for every single decision that the agents make is traceable back to a policy element. And we're giving customers like the logs of what the agents are doing too, as opposed to just kind of like keeping that behind the scenes. Right. And what that means for clients in, in TLDR, I guess is like much lower decisional rates, smarter reports effectively tuned to your

[00:14:43] policy, not to what we think your policy should be, what your policy is. And third, like dramatically lower cost and labor associated with adjudicating these reports. Um, so it's part of this wave of like moving from software to kind of manage service in a way in the world of social media screening. I think it's like, uh, uh, a shift that we're going to see across the marketplace as business users go

[00:15:09] from, this is not novel, but like interacting with software tools to, you know, engaging with vendors on outcomes, you know, so to speak. So anyway, that's it. It's called Fama Lens. It's live now. We're super excited about it. Like again, first time I'm talking about it. So it's new, uh, but yeah, Fama Lens, that's what we're, that's what we're working on policy-based screening, baby. So if a prospect or a customer doesn't have a policy, so I get, and I think the audience gets the

[00:15:36] policy thing. So, but if they don't have one, do you have a place for them to start? Do you have a library or something like that? So this whole world of engaging with business customers from a historically like software perspective is that you aren't just like selling a UI. No, you're not really selling a product. You're selling a genuine service. And so part of the

[00:16:03] onboarding is that we put together a team, like we resourced it proactively who essentially has, yeah, a adjudication policy building session that they take customers through. They record and transcribe that conversation. An agent builds a policy. We send the policy back to their customer. The customer then stamps it, sends it to legal, make sure everything tracks. And that process can take a little bit, you know, getting through the weirdest pieces of the organization.

[00:16:30] But dude, in the world of screening, I can't tell you not just in social media, but in the world of background screening more generally, if you ask five different people in your company, like what's our background screening policy, you might get five different answers. Like where do we draw the line? What do we say yes to? What do we say no to? And this sort of takes the governance and puts it front and center, which weirdly enough is more compliant than the old way of doing it, especially a lot of

[00:16:57] like principles of least intrusiveness and minimizing invasiveness of inflicting, like you were saying, insight in someone's personal life. Now, you know, uh, an end user is never actually going to see any of the things that aren't mapped to a specific policy element. So it's a very strange and new world we're living in, but one that to me is like super energized and see a lot of ways saying we help the customer create the policy. Cause to your point, like even one off the shelf is not going to be

[00:17:24] native and nuanced to you. No, it'd be cut and copied from somebody else's policy. But I like the fact, uh, that employment attorneys, the legal team is involved because they, especially with all of the stuff that's trackable that they can see, they can see it. And it probably makes a hell of a lot more sense to them now than it did before. Yes. Yes. Right. Oh, I love talking to you sometimes because you, you, you pick it up. Yeah. That not, I love talking to you

[00:17:53] all the time. Yeah. No, I got, you're right. Like, so let's just talk about how companies buy stuff. One person's like, oh my gosh, I love this thing. Like, let's do it. Then they, they're like, okay, let's go talk to legal. I'm going to talk to HR and employee relations. And they're like, wait, what is this thing? And this person's like playing telephone. Like, wait, what are you going to screen for? This is a bad idea. This is a terrible idea. Look at people's social media. We're going to see

[00:18:20] like their only fans account is theirs. It's not ours. Their only fans account is theirs and they can do whatever they want on that, on that little website. I have one. I have one by the way. I have an only fans account. I do. I have 66 subscribers and I give HR and recruiting advice. So I could have done a page. Yeah. I could have done a Patreon account and I'm like, yeah, not the same bit. I'm

[00:18:46] their, their verification process is world-class. Like, are you, you, you like they're, you know, like, so if you ever think not that you would, but if you ever, ever, if everyone would ever think is the person on the other end of that, is that a real human? They have to be like, it is fascinating. So yeah, I've got 66 people using only fans for business reasons, but oh,

[00:19:11] it's fantastic. I mean, that's a, that's a, that's a thing at the end of the day, you know, it's like, instead of playing this game of like enterprise GTM telephone, like you said, you just give someone a policy and it's like, this is what we're going to do. This is what, and it takes this idea and this concept of like, well, how are we going to interact with the product? What's the product going to do? What are the flags going to look like? What are the keywords going to look like? What's the,

[00:19:38] you know, look back, it's rules-based at a certain point. You've made it rules-based, which lawyers love because then they can, like, they can understand it because that's something that they've been taught. The law is rules-based. So it's either, it's either in the rule or it's not in the rule. And so it's, it's, it's a wonderful world for them that HR bumps up against all the time because they don't play and they don't use the same language. And so now I think that that's what's important is now they're not doing out doing something that's way

[00:20:08] out away from the business that might create risk or compliance issues, et cetera. They're doing something strictly within the business or what they care about and what's legal and it keeps them compliant, which, you know, HR does care about compliance, but lawyers, absolutely. It keeps them up at night. Totally. And, and that's sort of where I think all of this is going is into

[00:20:34] this removal of the translation layer of how software products engage with corporate. Yeah. I don't want to say corporate governance. It sounds too like buzzwordy and being corporate, but like it takes the things that you actually want to do that you're comfortable doing, which is like writing a word document, editing a word document saying like, this is what the software is going to do. Then of course there's a bunch of changes around or questions around like, what's the AI governance framework? Is there a kill switch there? We, we had to put like

[00:21:01] deterministic rule layers on top of the agentic layer, put a kill switch in place, active model drift. So part of that initial, this is a little inside baseball, but like part of the like initial pitch package for this product is you actually send the customer, the entire governance policy for your software to not theirs, but you send their policy, they drafted and how the AI is going to execute. Well, instead of it, we grew up in a world where it was black box.

[00:21:30] Yeah. And people would trust that, but also not trust it. And I think we've gotten out of outside of that and the companies that lean harder into, here you go. There's no black box. Here it is. This is what we do. This is how we do it. And if we have some problems, let's talk about those problems. Let's, let's get those things taken care of. It's not, I mean, again, the people that are afraid of transparency, it's been my experience at least, are the ones that have something to be afraid of.

[00:21:56] Yeah. It's, it's, it's, there's, there's a reason you're fighting your transparency and it has nothing to do with transparency. So I think I like the fact that you're putting that stuff forward and just most people won't ever read it or care. However, the fact that it's available to them is important that, you know, I did, I have this been available to me before because y'all would just run the thing and then, you know, candidates would show up and not show up and that type stuff. So

[00:22:26] I, I could see where they'd appreciate that more. I think in the next turn of how all of us vendors and like HR tech, I'm sure around here, we've got AI everything. Right. But I think vendors are going to win now if, because they're asking their customers to trust them with so much more of the workflow and the process data, not just the technology that

[00:22:53] supports it, but really running a lot of that. Like in our case, like we're, we're taking a business, you know, governance policy and we're screening against it. Like there's a lot of, you know, work that they used to do that we're now doing. And so I think in those environments, the vendors that are going to win are ones that proactively share how the sausage gets made, you know, like there were concerns and questions like, do we want to give our customers all the logs of what the

[00:23:18] agents are doing? Because yeah, they could see the prompting. They could see, you know, how they could reverse engineer it or something like that. And I was like, you know what? Like, yeah, if we're worried about, if we're worried about a company building a product like ours, Godspeed, let them, let them give them a lot, give them, give them the math, you know, give them, give them the money to do it. Like, good luck. You know what I mean? It's, it's one of these things

[00:23:41] that's like the, the further that you can go down this path and doing it well, you're entrenching value. You're not entrenching product differentiation. You're, you're helping them own the whole workflow. So the questions that were around probably two years ago and forward was the verification process of screening. We won't, not, not necessarily social screening, but, but, but of screening where it looked like a lot of the screening companies were going to become

[00:24:09] identity, I did, I did the verification, are you, you software? And oh, by the way, well, by the way, we'll run a background check or we'll do something else, but really what I got to do it first, I got to clear, are you really Tammy and is Tammy really in Portland, not North Korea? Like, so there was, it was a lot of, a lot of, a lot of discourse around the verification process. Um, and then the background screening took a kind of a backseat to that. I haven't seen as much of

[00:24:38] that. I'm excited to, to see at HR tech if, if that's kind of making a comeback or what, because I, I can see verification being important across all types of screening products, really across all, because there's always been fraud in recruiting and staffing, you know, the RUU stuff has always been there. It's not a, it's not a new thing. It's just kind of, it's just newly enabled. Um,

[00:25:03] so I'm, I'm interested to see if that becomes a thing. So the question is, do your customers ask you, do they, do they, do they ask you about, okay, we know that this is Tammy's Instagram account, which let's just say, um, do we, do we actually know that it's her account? Yeah. We look, the very, the world of verification, your point, just spend a second on that was probably the fastest growing, you know, part of background screening for the trailing 18 months,

[00:25:33] I would guess he probably will be into the next six to 12, uh, because that North Korean problem is real and employers to your point should be, are trying to run identity earlier in the process. So they don't waste time on an interview works, but it was at worst confidential information through the interview process that even if you catch one before they start, it's like, well, haven't we wasted a ton of time on this candidate already? And I'm like, gosh, we just shared our

[00:26:01] roadmap with this random North Korean person who, who knows what other data they have and what they enrich, uh, enrich, uh, enrich. Oh, by the way, payroll companies is a little bit outside of the things you'd care about, but payroll companies are learning about it during the onboarding process when they set up direct deposit. So when they set up direct deposit, they're like, wait a minute, Tammy's not in Portland.

[00:26:25] Tammy, Tammy's, Tammy's somewhere else. So it's like, okay, if they've gotten through the entire process, been onboarded and it's like, okay, what did you just expose yourself to? And, uh, and them too. So for sure. So, so on social media, we, we've always gotten that question of like, how do you know if this person is the right person, uh, as far as the right social media profile that they have?

[00:26:50] Interestingly, we've had a few cuts at like, uh, we've done some half week projects on like, we use person's social media profile to tell if they're real or not, or if it's a fake person. Right. Um, but you know, dude, like that space is so crowded now, like the identity management space is so crowded. I mean, this is like when the state department last year put out like guidance to

[00:27:15] employers and the background screening firms, like here's an emerging problem, like all of Silicon Valley and a lot of these like late stage, you know, PE backed software companies who wanted to add, you know, more kit, so to speak to what they offer just tripled down on that. So I got, I got some, you know, great advice from the founder of indeed once he was like, if you told me like, uh, or as we think about indeed, he was like, if you told me, you know, years ago, I'd all these people would

[00:27:42] come to me and be like, Oh, you guys are a great job, but what if you did like ATS? What if you did like this, if you did that? And he was like, I just kept it lasered in on job board. And that's what you're synonymous with at this point. So, you know, uh, despite all the ups and downs of that company, that was just great advice to me. If someone is just like, just stay focused on what you're doing and keep delivering products. So governance, governance is great with it when

[00:28:08] it's black and white. And you gave the example, I'm killing it with my boss versus I'm going to kill my boss. Um, I'm a, I'm a person that deals with dark humor, uh, and sarcasm. So sometimes I'll say something and it might be interpreted a certain way. Now you had some PhDs go through and make sure that that stuff is probably cleaned up, but do you get questions from prospects or customers about

[00:28:36] sarcasm is probably an easier one than dark humor, but, but, uh, you know, people can be dark online, but not necessarily dark. The, the quick answer to the question is I am continually blown away at the capability of what these models could do. We, I'll give you an example. We have a global customer that, um, had questions for us about operating in Brazil versus the States in Brazil. KKK is like,

[00:29:05] ha ha ha. Right. That's like how they, they write it in Northern Georgia. It is too. There you go. That's actually very, that's actually very funny. Uh, but, but, uh, you know, obviously that's not like the white nationalist, you know, moving into group. Right. And so when I saw the regional context that lens was able to capture of identifying the right, a KKK, recognizing the rest of

[00:29:34] it was in Portuguese and that there was a contextual, so there are tools that, that AI can do, but that shouldn't be heard or interpreted by you or anyone else listening as like absolute rules. Right. We're entering again, this paradigm of like how we interact with technology that's fundamentally different and service providers need to prep their customers that there will be things that, well, are not

[00:29:59] technically false positives and map to an example. I'll give you one, which is like, we have a, uh, a influencer customer, uh, meaning like they do, uh, they're not in front. They, they vet influencers to pair with a big brand. And because they market to kids, they can't have anything related to alcohol

[00:30:22] period, hard stock, anything related to alcohol. And we flagged this like ice cream cake on their behalf. And we were like ice cream cake. And this was during testing. It wasn't in production. We were like, is his rum in it? Yeah. Dude, this thing figured out that there was rum in this recipe. And we went to the customer and we were like, do you want this flagged? And they're like, yeah,

[00:30:49] we do like that. And so, you know, it's like, wow, that's good. That's good. Crazy. And so, you know, I think business users just have to kind of like get to this point of acceptance that just like a friend might not interpret a dark joke. Like the AI might not always, but damn, it's going to get close and it's going to be like understandable. Why? If he built all the anti-hallucin, like hallucination and other controls. And it's, and it's always getting

[00:31:18] smarter. Like all of AI, it's learning as it, as it goes as well, specifically for that customer. So as the customer says, yes, no, yes, no, yes, yes, yes, no, all that stuff. It's learning. Integrated memory is, is, is fascinating. Well, the, uh, the dark humor thing. So an example of that is, uh, like when you said KKK, I said, well, you know, in Northern Georgia, that's actually where

[00:31:41] the KKK originated. They found gold in, uh, on Cherokee land and that became the trail of tears. So they found gold. Oh yeah. They found gold on Cherokee land and basically said, you people have to leave by you people, uh, Cherokees. So Andrew Jackson, they went through the Supreme court. It actually went, um, all the way through state court, Cherokees won, went to the Supreme court,

[00:32:06] Cherokees won. And, uh, um, the president looked the other way and they were forcibly removed by the Kukhlaas clan. Of course. Well, it was the cavalry that removed them, but who encamped, who, who put them in encampments was actually the KKK. I never knew the KKK was involved in the trail of tears. It was the beginning of it. There would have been no trail of tears had they not found

[00:32:31] gold on the, on the, on their land. Yeah. So that that's borderline dark humor. It's also borderline history, but it's borderline dark humor. So some people will hear it. We'll go like, oh, that's dark. Not all the people in Northern Georgia. No, no, no, no, no. Actually at that particular moment, yeah. If you lived in Northern Georgia, you, you were a part of the clan. Yeah. And, and I think, you know, there's, there's something that we're getting to soon and

[00:32:57] anyone who's downloaded, like, have you played around with like Muse or Instinct or? Yeah. Muse. Yeah. I think with Muse, like we kind of get to the beginning of our understanding of technology and its fall ability, where I think when there was an idea that systems, that humans design should be perfect and there are companies that have to be perfect. Right. And when they're not, it's national news cloud. That's right. Cloudflare, Amazon web services, right? Those things go down and they're not perfect. If AWS goes down, we're all screwed.

[00:33:26] Right. You know, a hundred percent. I guess we'll be, I guess we're off today. Or when like, I remember I was in Chicago and like, uh, like AT&T was down in all of Chicago and I couldn't see where I was going or talk to anyone and find or I mean, that's a showstopper. Right. So I shouldn't say like all technology is like this, but I do think with consumer and business AI,

[00:33:50] I were understanding that the separation between humans and AI is in these sorts of dark jokes. Like you just also like, I know you, you and I have made way crazier jokes when we're not recording. Right. And so it's like, to me, I didn't skip a beat, you know, but yeah. And employers and candidates kind of have to do the same dance, which is like, all right. So I see why I said

[00:34:16] that this thing about KKK, but you know, you got it. But so the workflow for the, for the audience, the workflow is still the same. Um, it still ends up in your ATS. So you're compliant there. And then as you move forward in the process and whether or not you hire the person, all that stuff, all that stuff is there and it's well-documented. Got that. Where do you place, where do you place this now? It's always been something that's been interesting to me where,

[00:34:42] where do you place the assessments? Where do you place the test? Where do you place the screening? Where do you place this further back in the process or further front in the process? And, uh, and I know everyone's going to say it depends and that type stuff, but I'm like, yeah, it kind of depends, but it kind of doesn't like I, I, at the end of the day, I want quality candidates that fit with

[00:35:06] the hiring managers trying to, and I want all that other stuff out. And cause I, I only have so much time now. Hey, I might be giving me all my time back, so I might not need that, but it's kind of a bet on your view of your company and its composition. And, and in a way your, your view on humanity and, and all of that. But I, I, I think, you know, I'll speak to my company and where I think things could go

[00:35:33] with, with, with ours in particular, but in our world, like behavior matters, you know, my thinking is that the more the technology soaks up more work around the edges and, and you can turn to Claude or, uh, muse turns to Claude and does something on your behalf and does it for you. Right. And I, I believe more and more that humans are, are going to play an increasingly important part in decision-making,

[00:35:59] because now what rises to the top, but you can't fake is how you behave, how you treat other people, how you, uh, show up. Would you take a cookie from the cookie jar when your boss isn't in the room? Right. Do you, uh, make that comment about someone who's disabled online and belittle them, right? Knowing that you have an ADA accessible office and you serve, you know, that community, right? Those, those are the things that I believe. And obviously I'm long on Fama. I have a

[00:36:26] com strong commercial interest in saying what I'm saying, but like, I, I think you will see employers who today use us, uh, either at like two to three candidates per role or a conditional offer. So sort of the penultimate stage or on annual re-screening or, or biannual re-screening that they're doing as part of their existing framework and policies. So we don't call that, we don't call that monitoring anymore. Right. We does. It's just a, we, we call that a re-screening.

[00:36:52] So there is monitoring, which is like actual real time alerting, but on social media, we've never done like re-screening too. Even for, you know, Mr. Big Brother himself, that was too big brother for me. So, you know, I was like, we're not going to do real time monitoring. NSA called they'd like access to your database. Yeah. Got it. Yeah. Right. That's, but re-screening is a great, re-screening is a great way to frame it because it's that you hired them once you went through this whole process. They've there, they're, they've been there for

[00:37:20] five years, but you don't know really anything other than the work and the work product. Some companies do it every year. Some do it every three, some do it every five. Some people do it biannually that, you know, as, as attractive as the recurring revenue opportunity is for more frequent monitoring, we just don't see business users adopting it. So I think lens because of its ability to lock in with what your definition of what great looks like at your company.

[00:37:47] My hypothesis is that it gets start, uh, like run earlier, maybe shortlist. Yeah. For a position. That's I think for me, it's always been, I want to deal with less candidates that are more qualified. So whatever it takes for me to get to that place, I want to deal with 10, you know, it used to be with higher, uh, hiring managers and with recruiters. We had a thousand people apply to our jobs. Like,

[00:38:11] I don't, I don't care how many people apply. It's a vanity metric. Like we had a 20,000 people. I don't care. I just want the 10 viable, most viable candidates in front of my, uh, hiring manager. And these people have already been vetted. They've already been screened. They've already been all of that stuff so that they don't waste time because hiring managers, when we get right down to it, they're not taught how to hire generally speaking. And so they're, they, they, they, there's a whole

[00:38:40] lot of stuff that could mess up right there. But if you give them great candidates, there's less mistakes. Yeah, for sure. So well said, well said, I, I, uh, I love having a hard decision, three or four for a position and being like, I'd love if any of these would be, would be a fit. And what happens at the top of the, the top of the funnel, no one even like talks about it anymore. Honestly. No, no. Yeah. It's a high praise to a recruiter when a hiring manager is like,

[00:39:09] damn it, I'm pissed at you. It's like, why are you, why are you mad? It's like, I can't, I, I like all four candidates. I feel the exact same way. Yeah. Like that's high praise. Like, that's like, yep, I did my job. Cause then the recruiter's job is over. Then the recruiter's job is like, whatever happens from here, I got to get you four. I can get you four. One is the job, four is your, your exceeding. That's right. And, and some of those hiring managers I've seen,

[00:39:35] I've talked to recruiters that basically they'll go back to the CFO and get budget so that they can hire all four. Cool. Hey, done. Let's create a little growth space.