The Expert Network Behind Handshake AI’s Model Training w/ Garrett Lord & Mamoon Hamid

26:24Is this wrong? They need somebody with a specific set of expertise to basically look at whatever that model is producing within a given domain and then ask the expert. Yeah, well, it's a great, very simple story, which is like, you know, people talk about agents all the time, right? Like agents requires an ability to follow step-by-step instructions or, you know, fundamental idea behind agents is you understand like a reasoning and the reasoning choices need to be correct. So if you ask any one of the large frontier models questions and you ask it to break down its thinking of its step-by-step instructions, too oftentimes these models are not getting the answers correct.

27:02So when you think about a model, it's like you have compute, you have algorithms, and you have like data. Those are like the three things that matter with models. Everyone's applying a ton of compute. Everyone's constantly evolving algorithms and people are obviously chasing after one another in terms of algorithmic prowess. And then you have data and data is what really fuels the gains on what are called the post-training side of the house. post-training as human data. And what we do is, you know, the world's evolved from like generalists where you were predominantly leveraging like international skilled lower cost labor, taking advantage of an arbitrage between that to now these models have kind of sucked up the entirety of the entire corpus of the internet and every book and video.

27:44They've also sucked up and they've gotten good enough where like generalists are no longer needed. The models are actually better than generalists. And so what they need is they need experts. They need experts in accounting and law and medicine. They need experts in coding. They need experts in STEM domains like physics, math, chemistry. And designing a network that's focused on experts and treating those experts well and making it an amazing community experience with high retention where people can, you know, we're connecting to like badges on your profile and leaderboards by school. You know, we have access to millions of experts.

28:18So these experts, it's pretty cool. you can make like 60, 80, 100. You know, I think our average bill rate is over$100 an hour, like, you know,$125 an hour applying your expertise in the domain that you're studying. You know, you're getting a master's in accounting and you're getting paid$100 an hour to provide data and correct models. And yeah, we're doing really also wild things like, you know, audio, like a ton of audio work, like these models. I can't get, I won't get too far into it, but, you know, in the areas of trajectory, and tool use and text data and coding data. Like Handshake has, in our opinion, like one of the largest networks of experts that are captive and trust Handshake.

29:02And so we're trying to build a great experience for them to produce data for the labs. And can you just walk me through like a, how, like, okay, they find an expert on math or science. What is that expert actually doing? An average frontier lab, of which we work with six of them, most of the frontier labs, let's pick something like mathematics. So they all care about, and let's merge mathematics with educational design. So an analogous type of project that we would do for a lab is actually explaining and breaking down the decisions around how you would teach a particular math problem to a student.

29:43So you'd be applying mathematics experts who would actually agree on the step-by-step, like, here's how I would solve the problem. And you'd merge that with educational design PhD around the whether you like option A or option B more. And you might even get more specific, like you might provide like hints, or you might provide data around students in the Southeast prefer this educational design pattern, whereas students in the Northwest prefer this educational design pattern. So it's a public paper. So you go in, if you're a physics PhD, you break the model. you identify a reasoning error in a step-by-step instruction flow you correct the reasoning flow and then you provide the ground truth right answer and then they do things like you know pass at k1 pass at k3 pass at k10 which is like does the model because they're non-deterministic does the model get right one out of three times you know three out of three times get right one out of 10 times so we'll actually go in and provide distributions of data around how many times is the model passing this question?

30:42Here's what the right answer is. And that would be an example of like a physics PhD focusing on like their area of research, asking, you know, questions the model the model cannot do today and providing the actual right answers. Most of the data we produce, if you were to read it, like, you know, you could lock me in jail for like 30 days and ask me to solve one of these physics problems. Like, you can't even like, I can't even barely read the type of data that we create because, you know, it's really experts in PhDs. Yeah. And then like the goal is, OK, frontier model gets better as a result.

31:22And the next time somebody, Mamoun, like asks the model a question around the thing that you just asked it, it'll produce a generally accurate, a generally accurate response. Yeah, I mean, I think we can even make it simple with that, which is like, so model companies, they measure themselves. There's many ways to measure themselves. I'm just, this is just our, you know, a point of view that we hear regularly, which is like, there's like a model card. So if you look at like one of the latest models, let's pick like Llama 4, there'll be a model card. And the model card talks about its skillset in coding and MMLU.

31:56It has all these like what are called benchmarks or evals. Evals are essentially, you know, you try never to expose your model to the actual eval because then it will know the right answers. So it's like, it's held out. It's like, you're never trying to give that data into what the model's being trained on so that you actually understand whether it's improving in an area or not. And so what customers regularly come to us with and ask us for is data that helps improve a capability area of a model, targeting getting better on these evals or benchmarks. And so they're asking for, if you can imagine what they're focused on, it's like all the economic use cases of models.

32:34They want more professional domains. They want to be able to use tools. They want to be able to, you know, do trajectories to actually build, like solve problems. Like think about like investment, you know, where you're putting together like a total addressable market and then building a slide deck and then breaking down a model. Like they want to be able to stitch together many of the labs, like they had an entire flow, you know, agentically. And so we, we're providing data that they use to evaluate their model's performance and to train their models to make it better in certain capability areas.

33:07Like, so you go to the holiday party, you like start, like your brain starts going, Colin Mamoun. Like, how long did it take before you decided, oh, like, we're going to build this product? Like, we're going to go for this. Yeah, two funny stories there. one funny story was like, so I, you know, you can never pick the timing of when you get all energized about something. So I got super energized. I like, I got charged up. I like canceled like every, I mean, there's a couple of things you can do now that we have more resources that you couldn't do when you first started Handshake. Like one, you can hire expert firms like GLG and Tegas and Alpha Insights.

33:44I know their names because we hired them all. And you could be like, I want to spend 50 ,000 bucks on each one of these expert networks. Like sign me I first signed up to all the expert networks. I started basically going into my network and trying to meet with anyone and everyone that would talk about the space. I basically canceled basically everything. And I would sit on these expert calls. I wake up at like 5 o 'clock in the morning because it was 8 o 'clock East Coast. I was so energized. I would just go nonstop talking to researchers and product managers and people at companies that worked in the space and academics.

34:17like Tegas and these other companies are pretty liberal around who they'll bill you for on your time. So I started to, first I started talking to everyone and I started whittling it down and whittling it down and be like, I want to talk to, you know, I really built a point of view on how I want to talk to you. And then I was like, wow, we are so well positioned to build this business. Like I, and then I was upset with myself. I was like, how did I miss this? Like, this is, This is ridiculous. And my wife's like, you know, it's okay. It's okay. So then it was like Christmas. And so my wife has family.

34:52She has their home as a place in Arizona. And I just told the whole family, I was like, I am, this Christmas, I love all you guys, but like I am, I basically am like not going to be here. And so I remember, you remember I was talking, I would just fly out of Phoenix and I went to like New York and I went to like Seattle three times. I was like, just like chasing people. because one of my key things I think about building a company is you got to like hire great leaders who can help you build this business. And I recruited like three incredible folks who joined the team. And I also built this like elaborate spreadsheet of basically like on these GLG calls, I would ask like who I should hire.

35:27And I basically had like a point system that I was not that savvy, but just like, and I had this one guy who will remain unnamed. He'll know who he is, but he consistently came up at the very top. And there was another guy that came up at the very top. in the industry. And now the two top guys everyone thought I should hire, I have hired both of them now on the team, which is pretty awesome. And I did not take note with this one guy. I mean, almost lost a guitar on like a last minute plane ride. Like, you know, just like, you know, you got to chase, you got to go full force. Insane. Can, I do not want to divert too far from this because I have so many more questions, but the way you recruit is a full contact sport.

36:09Yeah. can you talk about it's gotta be right every founder's like this you're like not every founder yeah a little different yeah full contact yeah some people people are playing tennis yeah you're like the bruiser on the hockey team michigan yeah exactly like uh when okay when you like for example like be specific about these two ai researchers when you say i wouldn't let him like i wouldn't let him go like i wasn't going to take no for an answer what does that mean like you have a first call, they say, sorry, but I'm getting paid a gajillion bucks at wherever I am. Like, I'm, you know, I'm happy here.

36:46I'm working on this project. The normal stuff that you hear. Well, both are kind of iconic stories. I don't know. It's interesting. I mean, definitely interesting. So like one, uh, one of the guys, he had accepted another offer to join an early stage company. And this is like, I think it's like right around new year's, like right before new year's, like in between Christmas and new year's where like a lot of people aren't working. And he was the number one guy on the spreadsheet that I had been told on there. And I did an intro call with him, like introducing myself and, you know, talking about our business and why I thought we had the right to win.

37:20And he was pretty intrigued. He's like, well, can we meet, you know, I'd love to meet again. And like any, you know, good sales rep or founder, you like book the next meeting on the call, you know, you don't like push the email, the schedule. And he's like, well, I have a bunch of family stuff going on. Like I can meet tomorrow, like 11 o 'clock. I was like, cool. Like, go see tomorrow 11 o 'clock. So we talked again at 11 o 'clock. My whole Wolf family in Arizona was like, this kid is just weird. Like, diabolical. It's just a phone call. You're still, you're in Arizona and you're on a phone. And then I earned the right because I didn't want to be a little bit too aggressive and like fly out to him.

37:52But I like was trying to get the right to fly out to him because that's a little weird if you just say I'm going to fly out to you like right away. So, you can't look too desperate. So then, then at that 11 o 'clock meeting

38:05flew out to him. We ended up like taking over. Where'd you fly? I flew to San Jose and then drove up to him. And we ended up chatting and chatting. And he was supposed to start in like early January at this next company. And we spent, I think I spent like three or four days with him in person, brainstorming and building alignment and learning from each other on what we could build and what the assets were of Handshake and how much engagement we had. And he decided to join the team. So that's like, that was, that was. Sorry, you spent three or four days in San Jose? Yeah, in San Jose. Like you left Phoenix.

38:36Yeah. You left Arizona. Yeah, during the holidays. During the holidays. Well, I'd actually fly out in the morning and then fly back in the evening. And then I did, yeah, two trips. Yeah. Yeah. So by there. Full contact. I was also chasing a guy in Seattle. That was fun. And then I think one of the coolest ones, the even better story, iconic one, is this one guy, this other guy who lived in Qatar, gave me an opportunity. And I'd been hounding him to meet with him while he was in New York. for his 30th birthday party. And so I flew to New York to meet with him and he gave me a little bit of time at his 30th birthday party weekend.

39:13And we spent, ended up one hour straight into like a couple hours, you know, a whole morning. And then he canceled his birthday weekend plans with his friends and flew back with me and Mamoon cleared his calendar. I was like, I ain't coming back, red eye to San Francisco, like with, you know, the package inbound, like, can I meet with you?

39:35It was like partners meeting day and we got to meet. Thank you for this person. And now he's been good in both these, there's been several leaders, but both these two guys are incredibly instrumental in building our business. Yeah, full contact. And I'm curious, like, cause I think I was saying this a few weeks ago, Mumu, and like, I have not seen, like I have not seen many incumbent businesses. It's even weird to think of you as an incumbent, but like, you know, a business has been around for the last seven plus years, pre-AI era, that takes their base and then leverages it in such an advantageous way to like take advantage of this market right now.

40:17I have very few examples. You know, you've been building Handshake for a decade almost. Yeah. And I think of Handshake, you always had this goal to democratize access to opportunities for early talent. And this is not a departure from that goal. It is early talent. This happens to be master's, PhD students, and you're connecting them with frontier labs where there's real work for them to be done. But it is, you know, maybe a slight departure from the core selling software to Fortune 500 companies, which we still do. And we own that market as a company. And we own the career center at 99 point whatever percent of universities in the country and beyond.

41:08But I don't see it as too much of a departure. We just utilize that asset that we're building for the last decade to then build another business on top. So now compared to other companies that have done this, I just, in my mind, just going through this right now. I remember Roblox, very different company, had a good decade of just building the infrastructure to build the kind of games that you can build on Roblox. And it was not, I think for like a good decade, it was not like a thing. Like people, kids weren't going on Roblox and creating these, you know, characters and worlds and all that stuff.

41:45And it took a, but if it wasn't for everything that had been built in that prior decade, the Roblox that we know today wouldn't really make sense or exist. So that's a 20-year-old company, but only in the last decade has that hockey stick revenue adoption engagement happened. And then I think about like, you know, some of our own companies that, you know, we backed an e-commerce company called Amazon, yet enterprise value, you know, of AWS probably eclipses that of the e-commerce business. Or another one of our greats, Google, like, you know, search business, which is still an amazing business.

42:23but there are a couple of other businesses that have since then, whether it's Maps and then there's Proactivity and now Cloud. Some greats get to have Act 2, but it's after Act 1 is really working. And usually it's at a higher revenue base. It's not at like several hundred million. It's usually like a billion plus of revenue base from the core that gives you that wind in your sails to go do Act 2. And I'm not even sure this is act two for us, Garrett. I'm not sure if you fully agree, disagree on sort of the rationalization of where we are within the age of AI with Handshake. But it is, it's not common necessarily, but I don't think it's actually, this is like a big departure from the core mission of Handshake.

43:13I mean, think about it, like LinkedIn is an incredible business, right? They have, I think it's like five or six revenue lines over a billion dollars, right? They have marketing solutions, talent solutions, They have sales solutions. They have Linda and Learning. They have the JobSeeker Pro product. I mean, there are several business lines. So I think when you build an engaged audience, to your point, there are many ways to monetize it. And I think what I'm constantly encouraging our team to think about is how we really are connecting what we're doing in human data to our core marketplace. Because what we really want to do is be the best matching marketplace for people to find great jobs, right?

43:50And explore and discover and build like a whole network around that. And so what we're doing with this human data business is just fueling data that helps us get better at recruiting in our core marketplace. And maybe just for the audience, just the three-sided marketplace that we are today, just to maybe make that crystal clear. Yep. There's 1 ,600 universities in the country that use it, which is 92 % of the top 500 universities in the country. They're basically, if you're kind of older, listening to the call, if you know anyone in college, you could text them right now and there's a 0 % chance they haven't heard of Handshake.

44:23Like everyone's on Handshake in America. If you're 18, 19, 20, 21, trying to think about a first internship, first job, like you're on Handshake and you're on Handshake more than you're on LinkedIn. And we are growing that audience beyond college, like larger and larger every single year and aspire to like, you know, go bottoms up against these large incumbents as their trusted place they go to find internships, they go to find jobs. There's a million employers that recruit on the network, including 100 % of the Fortune 500. So every single Fortune 500 company, this is the primary way they hire students.

44:54There isn't a small business, medium-sized business, large business in America that recruits from college without most likely coming through Handshake because we power all of the schools. And yeah, the network's very, very strong in college. And we're growing in France and Germany and the UK as well. Can I, like, you made a comment of like, you're not sure this is Act 2. Do you think it's Act 2? I mean, it's definitely a totally new business unit inside of the company with a different set of capabilities and a different set of leaders. But we're really leveraging, and it's been very much like a zero to one in building this business.

45:26But a lot of what we're doing is monetizing the core asset and audience that, and we always dreamed of like monetizing and always thought we would be able to monetize many, many, there would be many ways to monetize this audience over the long term if we could get them engaged and get them to trust us and get them to use us regularly. It is also for PhDs and master's students for PhDs in particular, there weren't that many. I mean, it's actually like the community we're building with PhDs and professionals on the network now. Now we're acquiring like lawyers and accountants. Like it's really quite cool to build like professional community.

45:57If you were a PhD, like there wasn't really a place on the internet that you felt super welcome before. Yeah. Well, it's also kind of a no brainer. Like my mom was getting her PhD until she was 34 years old when she was like 33, 34. Like, and you told her from Iran, she could like be in the US and solving like, you know, like science problems for a hundred bucks an hour. She'd be like, is this a joke? Like all day, keeps your mind sharp. You get to meet other PhDs. Like that's kind of a no brainer. Yeah. Yeah. I'm sure you have some pretty amazing stories. Yeah. I think the, some, some of the best stories are like, uh, where students, when you think about it as a PhD student or almost all the people that do work on our platform though, people don't really realize this.

46:35They're like probably 90 plus percent of people doing work have already graduated. So So like people don't realize how large the alumni network is on Handshake. And yeah, the like a pretty awesome story is like just connecting like these dots, which is like, I want to learn more about AI in my area of research and actually apply it. Well, what's cooler than getting paid like$125 an hour to like, like learn complex prompt engineering, learn what models are really good at, learn where they hallucinate and fail in different situations, learn how to elicit failure. Like you're, what we're hearing from people is that they're actually bringing a lot of these skills then back into the area of expertise.

47:13And then connecting it to like employers, we're already doing this today where it's like companies want to be able to recruit these AI digitally, like the word right now is cracked, like software engineers, right? And so it's like, how can you actually be able to help articulate like who's really good at these tools? And then companies are asking us like, how can we upskill our young employees and employees with these goals? So we're really unlocking, I think, a lot of addressable opportunities around the business by participating at the forefront of like what labs are doing. it's amazing um garrett when the over the covet time frame things went crazy like hockey stick up into the right then zerp era ended that your tam started to saturate meaning like you were in most of the fortune 500 you're in most of the the schools right 490 or whatever whatever and so you start figuring out like all right like what you know like you want to keep growing the business and I remember you were like, there's a moment where the business is not growing like it used to be.

48:12And that was like very unnerving, right? You were like, you felt the ants in your pants to get back to the growth that we were at. And it's funny because you always talk about finding a second act and a third act, but like you didn't, sometimes you just kind of got to be in the game. And then all of a sudden serendipity hits you across the face at, in this case, our holiday party. And you're like, oh, here it is. Like, this is the thing. And you're kind of beating yourself up, but like sometimes you can't like plan for that. You can't like, I don't know. You just can't plan for these things, these moments in time.

48:42It's crazy. Yeah, I agree. I mean, I'm in this really cool organization I've learned a lot from and I've grown a ton from called YPO with other CEOs going through the exact same trials and tribulations, like trying to build companies. And there's this big proverb in a group, like you just can't quit. You just got to stay in the game. And yeah, I mean, building a company also for all the people that aspire to build companies or currently build companies that are listening, it's like it's, you know, my moon knows it's like, there's super highs and super lows. And, you know, you staying in the game, like you hopefully can create, you know, there's always ways.

49:17There's always ways to create opportunities. In our core business, like, we have less than like 3 ,000 companies that pay on the network. There's a million companies that use it. So like right now, our self-service business is adding a million dollars a week in addressing small businesses on the network. That business should be$400 million, like minimum,$500 million, just from small businesses, right? There's no pro products. Like there's no mid-sized products. So all we do is basically sell the largest companies in the world, great products. And we have opportunities to like level the playing field for medium and small businesses that can actually go out and be as competitive as like Google is in recruiting.

49:53So, you know, if we just drag out like single digit percentage, you know, conversion rates on a million companies using Handshake, there's still a ton of value to be built in the business. And then yet alone, like the post-college opportunity, the international opportunity, the opportunity when we think about like building this large matching marketplace globally. So I've, I remain like very optimistic, but yeah, we got punched in the face super hard when the economy like fell off, business started to decelerate. We had to, you know, it was, it was really hard. Um, I feel extremely like proud of all the, our team and the people that stuck through it and helped us like, you know, get back to that winning mentality internally.

50:31and now like confidence, the juju's high at Handshake. Like people are, people are excited. Well, dude, no kidding. Can I ask you on the, on the, how you organize the business internally when you have this new thing that's coming up and you said you do a zero to one, like it feels zero to one, like you're recruiting these researchers. Yeah. Almost as if it's a new startup. How do you, how do you orchestrate that? Same building? Are they the same resources? Are you sharing resources? I think we'll soon have a separate office just right now because it's growing like where I'm like people like 12 people are starting every Monday right so it's like and we'd love like 20 people to start every Monday so um I mean we just turned down a 20 million dollar project like yesterday morning because we just can't we can't deliver on it uh so the uh the way you operate I mean it's been very fun to like run back a second company inside of a company and you like learn you learn lessons I remember what you want to you know do the same way what you want to do differently I think the second time around like I've really been leading with like a lot of authenticity, like being, I really learned like authenticity matters so much.

51:35And like, you don't have to be perfect, but if you're like, you're, you know, a good person and you like, you're trying your best every day, like just like people give you the benefit of the doubt and just being more forward facing with the team. That was something I think I fell into like some patterns around when Hinshick was struggling of just being like too scripted and not authentic and too disconnected. So I've really learned a lot about that in turning around our core business around leading with authenticity that I brought from just day one in this new business, Handshake AI. You know, it's been pretty fun.

52:04I mean, last night we were at the office till 2.30 in the morning, which is like - Are you serious? Yeah, it's been pretty insane. Like lowest sleep scores my entire life for sure over the last several months. Lower than like the sleeping in McDonald's parking lot days? Well, I was like also 22 back then. So it's much easier. I'm like 35 now. It hits way differently getting like four hours of sleep at 35 than 22. And the team's like super excited about. So I don't know. I mean, to answer your question more specifically, it's like, you know, we have, you know, we've really kept things super flat in the new business.

52:36Everyone like rolls up to me, essentially. I have like direct exposure to like all the leaders running everything. I really am not trying to, in this new environment when it's small, you're like not the, you try not to be the boss. I like to say in meetings, like I'm just another really smart guy trying to participate with my opinion in the meeting and would love like to encourage active debate and like different points of view. I have some, we have superstar team that are like extraordinarily capable. I mean, this team that we built, like kudos to them and people are giving it so much effort.

53:03They're so passionate. They're trying so hard. We're just, you know, so the main thing I constantly am encouraging everyone saying is like kind of a lesson learned from my dad. It's like, you just never want to wonder what if, what's like we have this once in a lifetime opportunity, like just give it your all and try to lead every day and never wonder like, what if, right? Like give it your all. And then in the core business, we have incredible leaders that it's more scaled up. I mean, you set annual goals, you break it down into like what you accomplish quarterly, you break it down at the manager level to like what happens weekly.

53:35It's like, it's more systematized in how we track progress. Whereas, I mean, our demand tripled last month, right? So it was like, you can't have a set of like OKRs in this new business. So it's, hopefully that's helpful. Do you put any constraints around it? Meaning like, do they have unlimited access to capital? Like, do they, like, do you think about any of the constraints that a early company would have? Yeah. I mean, we, we have some like really constructive conversations at the board meeting. I mean, we want to operate at like a strong gross margin profile for this business. Um, and you know, the way it generally works right now in the industry is like, you know, you, you, you'll subsidize some of the early pilot work to basically break in the door and like prove yourself, single digit, like hundreds of thousands of dollars of like pilot work.

54:20Once you prove yourself and you're producing high quality data faster than anyone else, then you get an opportunity to like scale these projects and participate in more volume. And at that point, we're very transparent about our margin. People like working with us because we're just, they're researchers, like they're engineers. Like they prefer not to even be on the phone with us. Like, can we just like handle this via Slack, right? So if we can just make this as, you don't need to bring, it's also fun. It's like, I've, you know, I have traveled so much all around the world selling Handshake. And like, all I have to have is six customers.

54:50And they're all engineers like me who I don't like being sold to. And just on Slack. Yeah, just on Slack. It's like the best, right? It's like, do a good job and you will rip. And so we're doing a very good job. We don't over promise. We're super transparent on our margin profile. We, you know, we, and we just, you know, kind of beat and raise and like, you know, don't set too high of expectations. And the business right now is, you know, it's just flooding in. It's insane. I'm so happy for you. Thank you. I'm so happy for you. Thank you. And like, interesting, obviously, like just broader validation in the market that a comp, it's not the same comp, but a comp, Scale AI got bought for whatever it was, like$16 billion for - Half the company.

55:32For half the company. And to Meta, like, there's something here. I don't know what it is, but like you're fishing in an interesting pond right now. Yeah. And I would say just it's a testament to Garrett as a founder and, you know, the very trite founder mode. But it's true testament to just like when you have, one could say that the same circumstances, handshake as handshake with someone else at the helm would not have seized the opportunity the way you have. and that's just a testament to the grit, the resilience, the motor that drives you, Garrett. And I think maybe nine out of 10 times it probably wouldn't have happened if it wasn't for your ingenuity and your thought process and just your desire to go make it happen.

56:29Thank you. I totally agree. I totally agree. And you have to go inject energy into this business and only you can do that. Like the, that handshake energy that comes from the core that feeds into this different business. Like there's, that comes from you, you know? I'm stoked we got to do this. Thank you guys. Are you hiring? Like what? Like 20 a week. At midnight last night, we made a decision. We're hiring a franchise guy in New York. So if you're in New York, we're trying to build the whole team in San Francisco and build that like culture up and keep the kind of connectivity super tight. But there's just so much demand.

57:10We're now opening up Handshake AI in New York. So if you're in New York, we'd love to hire you. We're hiring dozens and dozens of people across engineering, operations, products. And yeah, it's a really, we would, we also have an incredible team that works in our core business, super talented people. We're hiring, you know, we're hiring at huge volumes there. So if you're passionate about this mission or interested in working at the complete frontier of where tech is headed, like we'd love you to apply. You can find us at joinhandshake.com. I can't help myself. I have one more question around hiring.

57:41Like this idea of headcount is going to get smaller, like that, that businesses are going to be built with way less people. I haven't, like, I haven't totally seen that. Like when there's an opportunity like this, it seems to me that most companies are like, all right, we're going to go take advantage of this opportunity and appropriately staff it. I don't know. It seems like you're doing the same. Yeah, I mean, we just get to focus people on different problems and hire less people. I think there's two things you can hold true at the same time. One, there hasn't been as much impact on our core company with AI as you'd expect.

58:20And B, it's also completely underhyped, and the impact over the time will be much larger than we can imagine. It also is these AI algorithms and neural networks in particular, like have been incredibly impactful in our core business. I mean, we have this like completely agentic interviewer where you, you know, Mamoun's taking it. You like, you interview with it and it grades you on a rubric. And we're like, you know, you can't hire 3 ,000 people in a month without using AI tools, right? So we're embedding all those tools in our core marketplace. Yeah, so I don't know if I'm really answering your question, but yeah, I think we'll see a significant impact to the number of people we hire at Handshake vis-a-vis these AI tools.

59:00And that's what we're hearing from companies in the network as well. Like what they really want. So there's, it's kind of fascinating. There's two schools of thought. One school is there will be no jobs for students in the future because AI will automate away all actual jobs. And the other school of thought is actually the companies want these like 8X, 10X, like digitally enabled productive employees. And, you know, take an example of like an engineer in our team I spoke with yesterday, like he had a PR up in a day and a half using cursor in our mobile code base. It's like on, nobody does that, like without these tools.

59:30And what we're hearing from our companies is that, yes, they're going to hire less people in certain roles out of college, but there's an incredible appetite for people like take social media marketers to like know how to use Dolly and like marketing analytics tools. Like there's, there's an opportunity to help these people really find great opportunities in companies that want, they want to hire one person. That's like the equivalent of eight people. and I think that's what we're hearing from our companies and maybe I'll just add that we're seeing our companies that are doing a lot more with less people or just shipping way more way faster what would take a year is happening in three months with the half the people and a key metric to just keep track of for all of us is this revenue per employee.

1:00:26And I think that's what's going through the roof is the efficiency or the amount of revenue being generated per employee is actually increasing at a rate that we've never seen before. Because what's happening is like you're building products, you're selling it for a lot more and have a lot less people building those products. And so I think that is really happening. I was actually just at a board meeting yesterday. Companies, Series A company built so much product already and direct question was hey you have 20 engineers uh how many engineers would have taken two years ago three years ago 80 engineers so you're seeing that truly flow through where tools like windsurf cursor and all the other supporting products that go along with building products that you can just do so much more with less.

1:01:22Totally. Guys, thank you. When you hear the word grit, what do you think of? I think my dad. Yeah. Yeah. Pretty, pretty gritty guy. Thanks for doing this, guys. Really fun. Yeah. That's it for now. If you liked the episode, please leave us a review or go back into the Archives, where we've done more than 200 episodes with some fantastic folks. This podcast is a Kleiner Perkins production, and I'm Juven. Thanks for listening.