Stop Hiring, Start Automating

Guest: Peter Guagenti, Founder and CEO of Everworker Host: Jake Aaron Villarreal

Jake Aaron Villarreal: I'm your host, Jake Aaron Villarreal, born and raised in Silicon Valley, and here to take you behind the scenes to share what it's like to be a startup founder — the journey they're on, the problems they face, the products they build, in an effort to make our lives better.

I'm excited to have with us today Peter Guagenti, founder and CEO of Everworker. Peter, welcome to the show.

Peter Guagenti: Thanks, Jake.

Jake Aaron Villarreal: Well, Peter, it's been a while since we've last spoken, but I guess before we jump in here, where are you joining us from today?

Peter Guagenti: I'm in Marin County. Silicon Valley, obviously, has a certain footprint that has extended north over the last 20 years. So I'm just north of the Golden Gate.

Jake Aaron Villarreal: I love that area. I used to live in the Marina, so I could look across the bay and go over there. I think the place was Sam's, if I recall — it was a nice little...

Peter Guagenti: Sam's in Tiburon. Everybody loves that spot.

Jake Aaron Villarreal: Yeah, nice little hookup there. But really cool — thanks for joining. We're going to talk about your company and what you're providing in the market, and the opportunities you see ahead. Before we do that, give us a little background on yourself — some of the experiences that shaped you and really got you into this space, and into technology, especially in this day and age. There's a lot of transformation going on.

Peter Guagenti: Yeah, for sure. I would describe myself first and foremost as a lifelong entrepreneur, and I think a lot of people say that, but I legit mean it. I grew up in New York City, always had hobbies and interests, and started having sort of side hustles in high school. I started my first couple of businesses back then, and was sort of smitten with this idea that I could actually build a business and make some money without having to work for someone.

I ended up going to school for sociology and communications. I was thinking I was going to work as a documentarian — I went to school working as a photojournalist. I just happened to go to university when the internet was just starting. I ended up dropping out of college in September of '95 to go and do some contract work building these new things called websites, for big brands. The benefit of growing up in New York was I had access to these larger marketing brands. The first contract project I ever did was for People magazine — I built People magazine's very first web property, when they were probably the most profitable magazine in the world at the time.

Jake Aaron Villarreal: Wow.

Peter Guagenti: They were something like that. Parlayed that into a consulting business, started my first agency only a few months later — I wasn't even, I don't even think I was 20 yet — when I started the first company, and ended up selling it a year and a half later to one of the big holding companies, one of the Omnicom holding company portfolio. And that just sent me down the path.

So my career's really had — I'd consider sort of three phases. In that era, I was very focused on digital transformation, helping companies really embrace this new technology, both from a product perspective, business perspective, marketing perspective. And then transitioned, just at the start of where SaaS started getting really interesting, into technology companies. Had a great career in core technology infrastructure, dev tools, that sort of thing. But then, as soon as AI started to emerge, that became my passion, and I really focused on what was coming with AI. I happened to spend a year at Google, actually running their global advertising teams, and got to see what was possible with algorithms then. And then one thing led to another and I ended up in AI.

Jake Aaron Villarreal: Yeah, I like the transition. I spent some time in New York City as well, and when I was there — what was the company that bought your company? It was Omnicom?

Peter Guagenti: Omnicom, yeah.

Jake Aaron Villarreal: So that was one of my clients, and actually — just a short story — I was in their office, I think it was up on the 20th floor, I forget the exact number, but this was back in 2001, something like that. And there was a brownout — elevators didn't work, so I actually had to walk the way down, and then I had to walk up like, I don't know, probably 40 streets, midday. That's my experience, or memory, of that organization — but really cool to see that you had that exit.

In terms of where things are today, I mean, every company we're talking to is doing something with AI — whether it's helping them generate revenue, or build a new service, or reduce head count, or improve the ability to do more with the same people. There's a lot of opportunity to transform. There's also a lot of different services and technologies out there that are helping, and it's hard to get above the noise of what actually is working, what's not working — everything's accelerating so quickly. Talk a little bit about what problem you saw in the market — you felt, "You know what, I can apply our services, or AI, or something to this problem and help solve it." Walk us through the space you're in.

Peter Guagenti: Yeah, well — I think people are starting to wrap their heads around this now, but AI is not just another technology, right? I think we spent 30 years adopting first client-server, and personal computer, and then the web and SaaS, and all these other things, and I think we had this pattern in our minds of what technology was. And deterministic, math-based applications made us more connected, they made us more powerful — they gave us access to information and access to capabilities we didn't have. But, probably at best, every one of those waves of innovation probably gave us, as individuals, a 10 percent productivity advantage, a 20 percent productivity advantage.

That's not what AI is, right? We had been hunting this concept of being able to have a true intelligence that lived alongside us — something that you could use to automate work at scale — being able to solve really complicated problems, problems that only humans were capable of doing, that sort of mathematics could not. And so the rise of the LLM really has changed how we think about work in general. And if you look at the early adopters, and what they're doing with it, what you're seeing is workers fundamentally changing how they approach their job — and teams fundamentally changing how they approach their job — to where you're seeing people doing 10x their productivity and throughput. You're seeing entire job functions be moved over to AI agents instead.

And so instead of it being the old-style org chart in a business, where you had leadership, and then you had middle managers, and you had individual contributors reporting to those middle managers — instead you're seeing functions shift to where everyone basically behaves like a leader. We saw this in software development first. It used to be that you'd fill out the bottom of a software engineering organization with very junior coders, many of them actually only lightly skilled — a lot of rejection in the code they were contributing. All that code is now generated by AI. You still need engineering managers, you still need to do real analysis and testing of that code, you still need to have a vision — you need the strategic thinking and the creativity — but now you have unlimited capacity with AI, and you have access to capabilities you just never had before.

And so Everworker was founded on a fundamental premise, which was: this is not a technology transformation. In fact, if anything, technology is actually relatively straightforward once they cracked the LLM — building an AI agent is actually a very straightforward process. This is more of a business transformation problem. And really, this is going to require reinvention of all of our jobs, and how we think about that. And the best people to reinvent those jobs are the people who know those jobs — the best people to solve that are the ones who really understand the process, how it works, and what they should be doing differently.

So Everworker was founded on this premise that we want to democratize the creation of these complex AI agents, or what we call AI workers. We want to be able to democratize that and put that in the hands of business operations professionals — people who really understood this stuff. And we've seen tremendous pickup because of that. We've been able to build a platform that takes all of the components you need for building one of these AI agents and fully integrates it into a single stack. You don't have to think about stitching together technology, or tools, or what's the latest and greatest of X in some part of the tech stack — it's just fully integrated, everything works together, it's all GUI-based, and you can solve for really complex problems.

And then, on top of that, we ended up building a catalog of agents for the most common business functions — areas where we already know the process well, we know what needs to get done. You think about, in marketing, things like content marketing — being able to go and generate content for various marketing purposes based on your messaging guide, your vision, and your thought leadership. Or in sales, we have AI SDRs that'll actually do your sales outreach for you, based on your strategy and your positioning. In support, it does things like read across all of your knowledge information to be able to actually help you respond to support tickets with the latest information available to you.

So we built a catalog, and that's also done really well for us — where if you're the CEO or COO of a company, you can come in and just say, "This is exactly what I need — these are the five use cases that are keeping me up at night, these are the things that are hurting in the business today, let's get them deployed." And then, being a platform, as opposed to just being a service you buy, you can fine-tune them all — because what we've discovered, once again, it shouldn't be a huge surprise to anybody who's operated a company — every company does a lot of the same jobs, but the 80/20 rule applies, where maybe 80 percent of that job is done the same between companies, but that 20 percent is different. More importantly, that 20 percent is probably 80 percent of your competitive advantage — that last 20 percent, that unique approach you have, is why you win.

So having a platform where we have these agents available, but they're all customizable and tunable, and you can tailor them to your specific approach to the work, means you get real value. And that was a really important thing to me when I spun up the business — I'm in technology not because I'm here to make money. I probably would have made more money if I worked as a banker. I'm in technology because I like solving customer problems. I like helping other companies build their businesses. And so I think the key with a lot of these tools and AI right now is we've got to get up to immediate time to value, where it can just be incorporated into how you work, and you start seeing that immediate impact on your business.

Jake Aaron Villarreal: Yeah, you make it sound really easy — I like it. It seems like there are some complexities to it if you're not technical, but sounds like you've created a platform that takes that out of the process. Talk about the type of customers you're focused on.

Peter Guagenti: We mostly focus on the mid-market. What we've discovered is, if you're SMB, I think the value of AI to you is probably going to be in personal productivity — you know, go grab a Claude, go grab a ChatGPT — it's about amplifying your day-to-day individual efforts. If you get into the large enterprise, they're kind of stuck in the old way of doing things. I don't think this is the right answer for them, but most of them are still trying to IT their way around every one of these problems — they're trying to code everything from scratch, they're trying to build all these agents from scratch. They look at a tool like ours, and — I can say this clearly — I pitched against a fully built-out Microsoft stack at a company that's a large enterprise, a big Microsoft customer. They could ship agents on our platform in a fifth of the time that their IT team could do using the Microsoft stack — but they fought it, because the IT team has a certain way of doing things, and frankly, they're trying to hold the rest of the business hostage. They'd rather do it that way, because then their jobs aren't at risk.

And so we just learned very quickly to stay away from large enterprise, because they get in their own way. And you probably saw the MIT study, or some — RAND did a great study — that was looking at these AI projects, and the MIT study was infamous because it said something like 85 percent of IT-led AI projects failed — they never even shipped, or they never got to value. So we just stay away from that.

What we discovered with the mid-market was two things. One is they've got all the complexity that the large enterprise has, but they don't have the time, or the money, or the patience to be able to work through those problems. More importantly — I think we're obviously all concerned about AI's impact on work, and on jobs, and all these other things — but there's actually a sub-story underneath all this that no one's talking about, which is: we're running into an issue with hiring inside a lot of these organizations. You've got boomers retiring — it's the largest generation in American history, they're all retiring — and it's hoovering up the next generation that's supposed to come and become leaders. But birth rates have been in decline for a while around most of the Western world. And so this is why inflation shot up the way it did — it's sort of a misunderstanding that it was COVID that shot up inflation. No, it was boomers starting to retire in 2018. And then COVID hit, and there was a reason to retire, because — I can't go to work, so why not retire?

And if you talk to most of these companies in the mid-market, they either can't hire for a lot of these job functions that they're trying to run and scale every day, or they can't afford to hire in that place, or they are hiring but these people are not capable. And so, you look at some of the functions that we can automate through AI, and it immediately solves for some of these hiring issues — so they can take their best talent and put their best talent on their hardest problems, their more interesting problems, as opposed to trying to just hire en masse to solve for use cases.

Jake Aaron Villarreal: Yeah. So you talked about a couple of the functions — you talked about sales, marketing. What other functions do your agents really focus on? You have use cases in a category of agents you've created — but what are you seeing as the most popular usage for your agents?

Peter Guagenti: Yeah, so because we're a platform, we see the whole long tail. I have agents in use that are doing very obscure back-office things for a business, where it's one of one — that's the critical task that business needed, and they couldn't get enough people to do it. For example, I'm inside a couple of software products, where we have agents actually powering the software products. One of them is for a SaaS spend optimization platform — so you can go in and put in all your contracts, what are you spending money on, and it helps you figure out what you're actually consuming, what you're not, how do you save money, that sort of thing. But to solve that problem, you literally have to physically read every contract — you have to know what that company bought, and contracts are ever-changing. So even somebody like Salesforce — their contract changes monthly or quarterly, and they change product names and all that stuff. So we have agents that will read those contracts and identify what you buy, how much did you pay, what's the term, all those things, just like a human would, and then puts that information in a place you can use.

So that's sort of the obscure end of it. But I share that with you because there's a lot of that — there's a lot of, "Hey, we ended up building a job function around this one thing because we couldn't solve it with software before, but now we can solve it with AI." And that's pretty transformative.

What we see, though, is most companies are going after a lot of the same problems. AI software development was the first to boom — partially because the LLMs were really good at writing code before they even understood English, right? It's a structured language, it's very understandable — so they were really powerful for writing software before they were powerful to do anything else. So that's not a surprise. But then, we've always had this issue — this hiring gap in software engineering. There's, yes, this conversation around a lot of junior developer jobs drying up — but they weren't even really software engineers. These were people who were copy-pasting code out of Reddit, they were going and grabbing stuff out of GitHub — it wasn't good code. So that was always an issue. If you were running a technology company or a big IT shop, you couldn't get enough skilled talent. So that area boomed really fast. We don't support too much of that.

But what that indicated for us was, okay, well, what's next, where that's a problem? Sales and marketing is number one — partially because they've always been understaffed. Sales and marketing has always needed more capacity than they could get. A lot of the strategic thinking is where the value is, but then you've got a lot of sweat equity every day. Building out a marketing campaign — you think of these big, famous marketing campaigns we've all heard of through our career — brilliant idea, but you don't realize then there's a thousand ad units that have to be created, and a bunch of copy that has to be written. It's just brute-force effort underneath that. Marketing loves using AI for that — things like content marketing use cases, SEO use cases, things like ad generation and ad redesign — you design one ad unit, but now I need it in 50 sizes.

Those are the kinds of agents that we deploy, and we see them used in sales — the pipeline generation stuff, the sales development. It's actually doing the customized communications — truly one-to-one communications to a lead, based on who they are and their behavior. That's something AI can give you the capacity to do that you'd need hundreds of SDRs if you wanted to do across some of these mid-market companies. Things like competitive intelligence — so you're getting on a call with a customer and actually doing all your call prep and your call briefing before you even start talking to a customer as a rep. We see HR use cases, support use cases. The other really big one is finance use cases. And you don't realize just how time-consuming and burdensome cash management is — accounts payable, accounts receivable, collections.

We work with a property management company — it's one of the bigger property management companies in the world. They have like 30 accountants who all they do is deal with the billing on both ends. And it's brainless work — it's not work they even want to be doing — but you need it, because there's no systematic way of solving that with software. But now, with LLMs, you can, because you can train an AI agent to do the same work that a junior accountant was doing — reading an invoice, matching it to a budget, understanding — you're reading the contract, making sure it's actually valid, and then going and approving the payment.

So those are the kinds of use cases. Now, we tend to focus more on these sort of core operational use cases. There's lots of other sexy stuff that's out there, happening in other categories — we like the boring stuff. The boring stuff is where — we've got a customer in a traditional business who's single-digit EBITDA, very thin margins, and they estimate that by deploying AI agents, they can double their EBITDA. Show me any technology that would let us do that, even in the last 30 years.

Jake Aaron Villarreal: Yeah, that's amazing. You know, how do you differentiate yourself versus other companies that are building? Because we're starting to see a lot more platforms come to the market claiming to be agent-focused, and easy to get in, to upgrade, to automate, to integrate with. Walk us through that.

Peter Guagenti: Honestly, there are a thousand ways of creating agents or buying agents today. And what we believe the differentiation here is — it's not in the technology, it's in the go-to-market. Who you serve, how you serve them, how you solve the problem. And so what we've really focused on, in terms of differentiation, first and foremost, is really just picking who we think we can be the greatest benefit to. And saying — these sort of boring, sub-billion-dollar companies — that's part of it. That was our differentiation, as we looked and said, we know these people need help — we know that they don't have the capacity to go hire a fleet of AI engineers to go and build this. So we're going to focus on them, we're going to focus on their problems.

And then we really differentiated ourselves with the agents themselves, by having these pre-packaged agents — I think about them almost like templates. They're ready to go out of the box, but you can tune them, you can change them, you can modify them. That became a big thing, because what we discovered when we first started selling the platform was there was a blank-canvas problem for a lot of customers, where they know the potential value of AI, but they don't understand AI well enough, nor have they even experienced it enough to know what the first project should be.

So being able to walk in and say, "Well, what are your pains today? Pipeline generation is a pain, or brand awareness is a pain," or — for a big retail company we worked with before, their physical locations was a pain, they had armies of recruiters reading through resumes — okay, let's focus on that. Let's go and drop in an agent that can support you there, and add immediate value to the business, and then that starts you down this path of a transformation to becoming an AI-first company. But every journey starts with the first step. And so we've really tried to focus on that first step, that's highly relevant to an individual customer, and really focusing on value.

And then, in our delivery, we really have this sort of interesting mix of platform plus expertise, that comes together to be able to solve these problems. So there's no over-promising and under-delivering. I've got a platform that could literally be used for any complex agent, but I'm giving you a starting point that's very consumable and very focused, so you can get started.

Jake Aaron Villarreal: Who cares about this product? When you're going to market, who's buying it on the inside — is it the founder, is it the CEO, is it the tech lead? Who are you having success with, selling this to?

Peter Guagenti: CXOs. It's directly with CXOs — which I actually think, look, it's a good and a bad thing. I think the CXOs are the ones feeling the pressure, they're the ones feeling the pain — they're the ones looking at the economy right now, and looking at the market right now, and saying, "I've got to adjust, just because there's downward pressure on me." I mean, I think we're not talking about the fact that we're sort of in a quiet recession right now, and a lot of these companies are struggling, where revenues are flat, and their costs are going up. I heard there might be an issue with access to oil right now — so if you're one of these boring, sub-billion-dollar companies, that's material for you.

So this is a conversation that typically is sold at the CXO level, because they're the ones who are worried about revenue, and margin, and profitability, and keeping the business alive. And so they're looking for any lever they can pull that'll get them there. What's really interesting about this is it's not happening with their lieutenants. I think there is a fear of AI and a fear of change that's happening in the layer just underneath them — and if you are listening to this podcast, and you are that person, that is a problem. And it's not just a problem for the short-term of the business, because you will end up being replaced by somebody who's AI-forward — you will. What was the old joke when AI came out? Your job likely won't be replaced by AI, but it will definitely be replaced by somebody who uses AI. And we're seeing that already.

Like, getting in, for example, on the sales and marketing side, getting into director level — "No, no, no, I built this, I built it my way, it works, it's totally fine" — and just sort of ignoring the data, ignoring what's possible. And it's to their own detriment. We're seeing some CROs even come in and be like, "Okay, well, we're going to ignore that person, we're going to put this in instead, and then I'm going to replace that person" — because I need a better mindset.

So you have this weird thing happening right now with AI, where it's like a sandwich — you've got CXOs all over it, really trying to figure out how to leverage it, how to reinvent their business processes to incorporate it. And then you have junior people stepping up and saying, "I'm using these tools every day, I'm using chat tools, I'm using dynamic generation tools, I'm using research tools." You talk to somebody who's in that sort of 25-to-30 range, early in a company, in an IC role, and they probably use ten different AI tools in a day. So I do think that there's an opportunity here for those people who have been sort of middle managers to become leaders, if they just start rethinking their own role and their own method — because they're the ones who have the expertise, actually. They're the ones who really understand how to execute this stuff best.

Jake Aaron Villarreal: Yeah. Is it possible for a single SDR to generate 150 meetings a month with AI? Is that something you think, with tools today, is possible? Or do I still need more people to actually manage these agents to get in front of customers?

Peter Guagenti: I have two SDR managers on my staff, and only one just started — so we're segmenting our business out a little bit. We've got one focused on one go-to-market motion, sales play, and another one focused on another go-to-market motion, sales play. And just with those two bodies, I'm going to generate over 400 qualified opportunities this quarter. This really is a 10x technology, and the sales development use case is an interesting one, because the SDR teams were always really challenging — you couldn't afford to hire the really sharp ones, because you couldn't keep them long. They graduated up into sales roles. All of my best SDRs, who were two or three times the performance of their peers, all ended up either — I promoted them into sales roles within 18 months, three, six months — or they got poached. So you always have this sort of weird capability constraint, where you're hiring a lot of people, but very few of them are actually sharp enough to really rock and roll in that position. And once they did, they kind of outgrew the position, or you just didn't have enough capacity to go and do it.

And the best practice for sales development has always been — empathy, standing, deep research, know who this person is, do really highly tailored, highly customized outreach. And if you're doing highly tailored, highly customized outreach, a rep is able to do maybe 20, 30, maybe at most 50 leads a day that they can prosecute. And then only, you know, 5 percent of those are going to convert into a meeting — the numbers add up really fast. I've had giant sales and marketing teams under me, and so you always have this issue.

With AI, you don't have any of those constraints — your constraint is based on compute capacity. And you're still doing the same coaching — my AI SDRs, for example, are using my messaging guides, my solution guides, my best practices, my competitive intelligence — they're coached on all the knowledge that a human would be coached on, all used in context in what they're doing. But a lead comes in, they prosecute it immediately — they don't wait, it's not — I don't have to wait till 9:00 a.m. Pacific the next day for the lead to get prosecuted, it just happens automatically. They're building completely custom sequences — those sequences are completely tailored, based on past performance. They're relentless in their follow-up — they just do — they follow up exactly against the playbook that you wanted.

And so what you end up with now is a great example of where an AI agent is not just more efficient in a role — it's actually more effective at the role, and it's more creative in the role than the human is. And so, yeah, the numbers are staggering. I have one customer who's using my AI SDRs — an open-source company — and they had 46,000 untouched leads that they were looking at, and said, "Well, maybe let's go and try to — let's go deep dive in those again, let's reach out to them again." And they processed 46,000 leads across a plethora of companies — there was some overlap in companies — prosecuted all 46,000 leads in less than two months.

Jake Aaron Villarreal: Wow.

Peter Guagenti: Standard capacity for email, because your constraint is actually how many emails you can send in any given day, as a domain. So they prosecuted over 46,000 leads in about two and a half months, and generated 3 million in pipeline.

Jake Aaron Villarreal: Wow.

Peter Guagenti: That just wouldn't have existed otherwise, because they didn't have —

Jake Aaron Villarreal: I love that. You know, you look at companies like Accenture and sort of big conglomerates that are out there — they've built their platforms, they're going to the markets, they're having these massive opportunities with hundreds of millions of dollars being spent to get up and running with use cases. But it sounds like you actually have a more nimble process, platform, and strategy. How do you compete with them — and do you? Maybe you don't, I don't know — I'm just asking the question.

Peter Guagenti: I don't, because actually, when I see companies — even in the mid-market — talking to McKinsey and Accenture and these guys, they're irrational actors. I'll give you a real example. I was talking to a company in the Netherlands — single use case, very focused. So we came in and said, "Great, let's focus on that use case, you don't have to do some giant, overblown AI strategy." Anybody who's building the AI strategy today — that thing's going to be a presentation that collects dust within six weeks, because the world is changing so — like, trying to do some massive overblown AI strategy is a waste of time and money. But let's crawl, walk, run — let's grab some use cases, let's start iterating you towards being an AI-first company.

And we came in, and because of our services-plus-software play, the very first couple of use cases we had was sub-$100,000. It would have been the easiest thing in the world to drop in the platform and a couple of pre-baked agents, some other things that need to be tuned, for less than 100 grand, and for about eight to ten weeks, depending upon how they wanted to move — we could have shipped the agent for them. Instead, they went with a competitor — we had to sort of pull teeth to understand who the competitor was. They went with one of the big consulting firms, who's literally going to charge them two and a half million dollars for the same project.

Jake Aaron Villarreal: Mhm.

Peter Guagenti: And when I talked to the CEO about it, it literally just came down to — we joked in AI, we went from FOMO to FOFO. So we went from fear of missing out, where everybody was all in and trying everything and jumping in with both feet — to where now they're kind of afraid and uncertain, so now they have a fear of messing up. Maybe not FOFO, FOMO — but we use FOFO internally. And so they're having these irrational responses.

So this idea of, "We're going to spend a million dollars with Accenture, and they're going to build an AI strategy for the business, and they're going to do all of this stuff, and then on top of that they're going to come in and we're going to do a couple of these agent builds, and we're going to build this long pilot and prototype, and we're going to spend a couple million dollars on it" — you're like, what — what are you doing? Like, why are you doing this? And by the way, as somebody who's been operating in AI for eight years, who's been at the forefront of this, I can tell you that no company like Accenture, at scale, understands AI at that level. You're going to end up — somebody sold you who's brilliant about AI, and you're going to have the same thing that they've always done. You're going to have some recent college grad who's sitting there doing research and reading other people's materials and building a strategy for you that's based on stuff you could have gotten off the internet. You could have fed the same response into ChatGPT and gotten the same strategy that they're going to give you — because what do you think they're going to do? Do the same thing, yeah.

So I think it's insane — personally, I think it's insane. And I say this as an Accenture alum — I was a strategy leader, I ran a practice at Accenture. So I know what works and what doesn't work in those organizations. And they're in a weird place, because I think AI is reducing the cost of access to real intelligence, to true superintelligence. And so the long-term prospects of consulting firms who sold expertise — like, how do you keep charging 400, 500, 600 bucks an hour for expertise, when I can probably get 80 percent of it just by knowing how to talk to Claude?

Jake Aaron Villarreal: What do you think the contraction's going to be like for a company like Accenture in 5, 10 years?

Peter Guagenti: They bloomed really fast. I was at Accenture now 20 years ago, and it was pre them getting into digital transformation — I was one of the first hires that was really focused on helping them get into digital transformation. And that was also when they started outsourcing — outsourcing was all about software dev, it was sort of software eating the world. And when I was there, if I recall correctly, there were only like 100,000 people — now that business is like 800,000 people.

Jake Aaron Villarreal: Right.

Peter Guagenti: I think as quick as it exploded, it is — it will shrink just as quickly. That's my belief. And I can say this also, having talked to CEOs who are more enlightened, who are saying, "Look, we've been outsourcing to Accenture for all this time, we want them to shift to AI for greater efficiency." And of course they're saying, "Yes, we'll do that with you, we'll start layering in AI practices in how we're doing it" — but they're not dropping their price.

Jake Aaron Villarreal: Mhm.

Peter Guagenti: That's not sustainable. Even the outsourcing stuff — I guarantee you I could find teams in Eastern Europe of PhD-level computer scientists who are using AI for software development, who could probably match the volume that's coming out of India, under Cap, and Wipro, and Tata, and Accenture. And they can match the throughput for probably a hundredth of the price, and probably twice the quality.

Jake Aaron Villarreal: It's amazing. Let's talk about people a little bit. You have a company now — how big are you today as an organization?

Peter Guagenti: We're small. I mean, we're AI-first ourselves — so we're only 48 people.

Jake Aaron Villarreal: Got it. And as you see the company today — you have to have bright people, it's always kind of hard to hire. What's worked for you in terms of making sure — in this new world, which, for a lot of what we see, and we're in the people business, helping companies find people and people find companies — the whole view of what someone does when they join a company is different. They have to be AI literate, they have to come to the table and be very end-to-end system thinking, whether it's building a product or bringing emotion to the business. They have to be able to have demonstrated it in the past, and what they could deliver based on outcomes that could help move the business forward when they join. We've had to transform our business, too, and it's no longer what skills do you have and how much experience — it's, you know, what can you actually get done with AI and leverage it when you join a company, and measure that.

Peter Guagenti: Yeah. So I hire for three core characteristics today that are different than I would have hired pre-AI. When I think about it, I've built a lot of businesses, I've built some pretty big teams, I've done the sort of zero-to-100-million run a few times in companies over my career. And I've always had sort of core characteristics I hired for. I've always been one of those people — I never hire for experience, I always hire for capability and for culture, and it's served me incredibly well. I can look back now, 30 years into a career, and I can't even hold count on fingers and toes any longer of people who I gave their first job, who are now executives, or even CXOs. And so I'm really proud of that, and I think being able to spot talent is one of the things I've always really leaned on in my career.

I don't think that changes — I think you still need to spot talent. But I think the core characteristics of people who are going to thrive in an AI-first world — I look for three characteristics. First and foremost is they have to have a passion for the function — they do this job because they love it, this is the work they want to be doing, they're passionate about it. I'm a passionate seller, I'm a passionate marketer. I don't just do finance because I happen to get an accounting degree — I do finance because I love orchestrating a business from inside of the financial operations. Those are the kinds of people I look for. And the reason why I look for that is because, no matter what happens and how the work gets done, if they are passionate about the work, then they will always find the most successful way to execute in that function. So in changing times, that works.

And I'll give you an example where it didn't work, even in marketing — the rules of marketing, and what worked and what didn't work, at its most aggressive, was changing about every three years, where the patterns that worked just stopped working, and new patterns emerged that did work — for building a brand, for building pipeline, for reaching customers, all this stuff. So I saw this even before AI, where you'd hire people who just did marketing because they fell into the role, and they just really repeated other people's patterns, and then they would fail — they would repeat the pattern until it stopped working, and then they'd keep repeating the pattern long after it stopped working, and just waste your money.

And so that passion for the role is really key, but it also leads into the second characteristic, which is — I hire first-principle thinkers. I don't hire people who copycat, I don't hire people who just look for an answer that's an easy button somebody else gave them. They are really thinking about this stuff deeply and saying, "Well, what's the right solution to this problem, what's the right approach to this work?" And they're open — I mean, you hire people who are first-principles thinkers, God, they will find the innovation everywhere, they will unlock things, because they're not stuck in a specific way of thinking — they're always trying to figure out what's the underlying root cause on this. And I'm that way — I read research and analysis and all this stuff, and I start asking the five whys in my head, and I can usually get underneath what's actually happening in a system very quickly, just by trying to unpack it and understand it. So we look for that.

And the third is being AI-passionate as well. You don't even have to actually be that skilled. I've hired some people into roles in the last year and a half who had dabbled with AI — they were heavy ChatGPT users, or they were playing with Claude Code, or they had started using some of the agents — a lot of marketers who were playing with Figma and some of these other tools — but didn't really yet have a full, comprehensive understanding of how AI actually works. I've been in the LLMs since they were born, so I have a very comprehensive understanding of how the technology works. I've also been building agents for four years, four and a half years now, long before people even understood what that was. So I have a really strong understanding of how this technology actually works and when it works well.

So you can teach that — that's not something you just have to come in with, you can teach that, you can learn that. But you do have to have people who are like, "Look, AI is a fundamentally new way of working, it's a better way of working, so I'm going to pursue that." And by the way, long enough in my career now where I've seen some patterns before — this is digital transformation all over again, this is all the same things that were true in digital transformation. I sat in boardrooms of Fortune 50 companies who stuck their head in the sand when digital was coming, and those businesses are gone. And all of those same three things I just said — they needed in leadership then, it just was a different world. And so every massive wave of technology, going back to the steam engine, has included a component where we, as the people who are now embracing that technology, need to really rethink our assumptions — fundamentally rethink our assumptions. And the core characteristics of the people who are successful rethinking their assumptions are always the same.

Jake Aaron Villarreal: You know, it's interesting that you say that, and how you've rethought how you go about finding people and hiring them. For founders out there that are going through the same process, and maybe some challenges — what's worked for you in qualifying that type of talent? Because it's one thing to say, "I need a first-principle thinker," but in the interview process, how do you calibrate that, how do you really filter through that?

Peter Guagenti: So a couple of things you can do. First and foremost, you have to be evidence-based in your approach — it can't be question-based. I have a bias towards things like exercises, discussions, presentations — more of that sort of thing, not just Q&A. And by the way, it amazes me how many times I still talk to founders and I hear some variant of, "Oh, it's somebody I'd really love to have a—" I really, honestly — I'm here to make money, not make friends. I don't need somebody to go have beers with — I have friends for that. But that's nice, I mean, you want cultural alignment, and you want that — but I want people who are really powerful and creative and capable and strategic thinkers and all these other things. And frankly, I'd probably be better off if I don't want to have a beer with them, because that means they'll probably challenge me, which is great — that's a good thing, right? If they're stretching you, that's a good thing.

So I think really being exercise-based, and really grounding it in — how do you prove that they have that characteristic — I think is key. The other piece of this is rubric-based hiring. So when I create a role, I always do the same thing, which is — what are the core characteristics I'm looking for, and what are the core capabilities I'm looking for in that person? And then, how do I find it — what do I need to see to understand — and I start with that. I don't start with a job description — a job description can come out of that. But I typically start with, like, what's the job to be done when they're on the ground with me, what's the gap? And, by the way, that's even how I build my organizations — I mean, I have patterns in my mind of what works for certain work structures, but I don't tend to think that way. Instead, I try to look at, what am I trying to achieve in the near, medium, and long term — near and medium's usually what you're hiring for. And then, where are my gaps in my business — gaps in skills, gaps in capability, gaps in knowledge, where are the holes with the people I have. And then I try to hire people who close those gaps for me in a thoughtful way, and then I construct my organization around my people, not the other way around. And that's always served me really well.

And I think, in the face of AI, the Altmans of the world, and the Amodeis of the world, will tell you, "Oh, all the rules have changed, you don't need that anymore, forget it, it's just AI's going to do all the things." I will say that AI is not creative, AI is not strategic, AI does not know how to break patterns to find success. It's incredibly good at repeating patterns that are already known, and it's incredibly good at processing information incredibly quickly to get you what you need — but I can tell you, brilliant people leveraging AI are much more successful than mediocre people leveraging AI.

Jake Aaron Villarreal: Yeah.

Peter Guagenti: So it is still people-first — it is absolutely still people-first in everything we do.

Jake Aaron Villarreal: Yeah, I like that. You know, it's interesting, because we're in the people business, helping companies all the time find organizations, or people find organizations, and vice versa. And one thing that we had to change was — it used to be, as a recruitment company, we get a job req, we look at it, we go in the market, we find the skills and the experience, we bring them to the company, and they hire them, and we kind of move on — that's just recruiting. But what we changed was understanding the impact on the business for why that person would be hired, and then understanding how it moves the business forward. Because if you find that, then you're actually helping build the business, grow it, scale it, whatever you want it to do.

And so what we started to do is say, "If you really want this business to change in a certain way, talk to us about how this person is going to help improve success for you in the first 30, 60, and 90 days." So now we know the function they're going to do, we also know the impact they're going to have for you, and we can then qualify candidates differently — because now we're not looking at their core skills, we're saying, "Here's what you have to do to be successful joining this company, and if you believe you can do those things, we'll make the introduction." Because now you know the function, and you know the expectation.

And then we took it a step further and said to our customers, "You pay us when they're successful in those 30, 60, and 90 days." So we're more aligned on what you need, but also on when the success happens — versus you spend the money, you get the person, you pay us a fee, you move on. The risk really starts when they start working for you, not before. And so we reversed that, and we've seen a ton of success from that model, because we're aligned more with what the company wants, what they need, what the candidate needs to do to prove they can do the job — and we also win if they succeed, we also lose when they don't succeed, we don't get paid. So it's just a different model, and I think a lot of technologies with AI, in particular, are kind of moving to these various different models — but it's all becoming more outcome-based. I'm sure, maybe some of how you structure your agreements — I don't know it — but a part of it is, yeah, you pay us for the work we're going to do, but maybe some others we're seeing are based on the outcomes you need — we'll deliver, we're also going to get paid, maybe certain amounts, as we succeed for you.

Peter Guagenti: Yeah, that's — it's kind of — I love that, Jake, I love that. And for us, we are value-priced — so we price the access to the platform based on the value that we're delivering. I typically want somewhere between 5 and 15 percent of the incremental revenue, or the cost savings, that I'm generating for you. And we're very flexible on pricing, tied to the use case. And I think that's right — I mean, for me, I was drawn to that as a product owner and a buyer, because of what you said about your process, which is — now you're aligned, now your values are aligned, your impacts are aligned. And I love that you're doing that with your people, because I'll tell you, I've been on both sides of this — I've been an entrepreneur and had my own companies, I've been an intrapreneur and hired in as an executive to scale other companies or fix other companies — I've done a lot of fixing in my career, I actually parachuted in to fix businesses that weren't doing so well. And I'll tell you, the first question I always ask is, "Why this role, and why now?" Like, why do you really want this? And then the 30-60-90 — I ask that question directly. What do you want this person in this role to achieve in the first 30 days, 60 days, 90 days?

And what you find, when you hire somebody who's not going to make it, it's your fault — it's not their fault, it's 100 percent your fault. You either didn't set expectations correctly, or you didn't evaluate them effectively. And so I never want to be in that position as a hiring manager, and I never want to be there as an employee — never. Because nobody wants to take a job and then feel like they can't succeed — no one wants that. And so I think aligning yourself as a recruiter in that mindset is really powerful. But it requires something interesting here, which is it requires what I said — it requires a rubric, it requires a real, thorough understanding of what you're looking for. And I can't tell you how many times I've worked with hiring managers where they effectively say something along the lines of, "I'll know it when I see it."

Jake Aaron Villarreal: Yeah, wrong answer.

Peter Guagenti: That will never work. That will never work.

Jake Aaron Villarreal: Well, what's interesting, what we found, especially in this world of AI now, is that companies have to think differently about the type of candidate and talent they need. We're talking about AI literacy, and leveraging tools that can come in and help you do the job, and being passionate about leaning into that. But what we've also found is that companies — and maybe more hiring managers, not maybe the founders, but the hiring teams — they haven't really clearly thought through what they've defined as success. So they've got their old-school mindset of, "Here's what I need, here's kind of the skills I need to show, and here's what they can do for us" — and that's great. But when you really push and press — well, what does that success look like in 30 days? And by the way, can you actually, with the inside chaotic process you have potentially with some companies — is that really fair, can you actually expect that person coming into your organization to do that for you?

Let me give you a quick story. We're working with a company right now — they're hiring a bunch of AI engineers. We sat down with them, they were very clear on what they needed. They said, "In the first 30 days, this engineer has to ship a feature. At 60 days, they had to automate workflows. And by day 90, they had to either accelerate how we develop, or reduce our head count or cost in some fashion" — as an end-to-end engineer. Okay, great. And when we knew that, the questions we were asking weren't "What have you built, how long have you done it, where are you coming from" — it was, "Here's what you need to do to succeed — have you done this today, or in the past, can you prove it? And if you can, do you believe you can do this in the next 90 days? Because if you can, we'll make an introduction. If you don't, we don't want to waste their time or yours" — because if you don't win, we don't win, they don't win, nobody wins. And then the conversation's more aligned with what they need, but also the impact on the business growth for that organization, too. So it's a very fun way to be part of this transformation, even though we're not engineers — we're helping find the right people.

Peter Guagenti: I love that, I love that actually — because there's also something in what you just said, which was, now you can also manage expectations with your hiring manager. Because what you just described are probably the most valued, and therefore the most expensive, AI engineers you're going to find. So did you price accordingly, are you planning to hire from the right places accordingly? You're not just — you're not title-shopping anymore, if that's what you're looking for — you're not company-shopping anymore, which is what a lot of terrible recruiters actually do.

I think there's an opportunity for AI in recruiting to solve for some of this too, because if you can get a really good rubric — every recruiter falls back on the same — and I mean these in-house recruiters, they all fall back on the same mistake, where they just start looking at titles and companies, and they don't even bother to unpack who the person actually is and how they approach the role. And if you're interviewing for somebody who's experienced, they've put thought into the description of the work they did. Even just going on LinkedIn — read the descriptions for people who are good at their job, and you're like, "Wow, this was amazing work you did here." But what inevitably happens is you just end up with that scanning. And I've even seen the AI tools that people are using making the same mistake, where they're ranking based on qualifications as opposed to characteristics.

And it's really interesting, because I think you lose — you lose gold — you lose people who are probably incredible and a much better fit for you, by doing that, as opposed to the sort of process that most people follow. And I feel like the best recruiters I've ever worked with — the best folks, like yourself, I've worked with — they scan, using that, to find people who are probably in the right wheelhouse, but they go deep on the individual. They're like, "Who is this person, what are they capable of, what have they done, what can they do, what happens if you unleash them on this problem, how effective will they be?" So you're asking those questions, and I feel like AI can do that at scale now.

And I'll give you an interesting study I read. There was a firm out of — I think it was Indonesia, somewhere in the Pacific — who was hiring for retail, and they ended up shifting their recruiting process to more of an AI-led approach, where they had questionnaires that they were asking people to fill out. They were actually evaluating every word, everything they said, to try to find the best fit for what it was. And then you could opt in to actually doing a screening conversation with AI instead of with a human — you didn't have to, but it was very upfront, it said, "Would you take a screening conversation?" And after doing this for about six months, they identified that the AI was better at finding the right talent — it actually had higher retention and performance rates. So on both measures — for the employer and for the employee — it was better, and they actually were happier with the recruiting process.

And the reason why — the thing that I think folks like you and I have always known as hiring leaders — is there's no bias, AI has no bias. If you coach it on what it's actually looking for, it's not going to be biased in its questions, or its response to you. You just have to give it the right coaching, and then it doesn't get tired. It's not going to shortchange the process because, "Oh, it's Friday at 2:00 p.m. and I really want to go home to my kids, and I've got three more interviews today, I don't really—" and they're checked out, they're not even paying attention when they're on the call.

Jake Aaron Villarreal: Yeah.

Peter Guagenti: And by the way, that doesn't just happen in retail — I have that experience as an executive interviewing for jobs. Sat on a call and I'm like, "This person is not even paying attention to me, they're totally somewhere else right now."

Jake Aaron Villarreal: Yeah, it's — I mean, what you're talking about is in line — I mean, we have — we can have a whole other discussion about what we're building internally — but yeah, I mean, we see the opportunity with AI, we see where agents could be a real benefit, we also see where human-in-the-loop is required. It's a people business, so there is the judgment, there is the assessment, there is some of the gut — what is my gut telling me. But at the end — you know, if agents don't get tired, and they could actually do voice AI interviewing — and maybe it's more for applicants, not necessarily so much for passive candidates that might not leave a Google — an interview on some system, because they don't know where it's going to show up, and if their boss is going to see something — but for more applicant-bound interviews, there could be an opportunity, but also within the cycle, there could be ways AI can help.

Peter Guagenti: Yeah, agreed, agreed.

Jake Aaron Villarreal: So, in terms of — I always like to leave a little space at the end for — if you are growing, if you are hiring, share what you're looking for, maybe if somebody will find you, or talk about that a little bit.

Peter Guagenti: Look, we're — I mean, we're growing, we're heavily focused on these very vertical-focused agents right now. We built a really powerful platform. We're always hiring, but we really want expertise in these areas — that's probably the biggest hiring area for me next, is more product leadership, who really understand their functions as subject matter experts. So if you are somebody who is a product person, who's got a passion around some core part of business operations — sales, marketing, finance — I want to know you, I want to know you.

Actually, and by the way, a little bit of a plug — Everworker, one of the founders of Everworker is a gentleman by the name of Ramin Timurshev. Ramin was the founder of Veeam — he is a bootstrap billionaire. This is a guy who literally came here from the Soviet Union in the early '90s, when the Iron Curtain fell, and built a life for himself here, and bootstrapped two companies and sold both of them — incredible, incredible guy. He's also invested — he's my co-founder and my primary investor, but he's also an investor in another half a dozen AI companies that he's incubated, and probably a dozen outside of that that he's just funded. So if you're great and you're passionate about these areas — not just for me, I can point you into one of the other businesses if we're not a fit for one of those roles. So yeah, we're always looking for folks.

And then I'll do something a little different with the ask — I'll flip it around. I'm also super passionate about helping young entrepreneurs. I don't have a lot of bandwidth for advisory work, but if you're listening to this, and you think I can be helpful to you, find me on LinkedIn, reach out — I love helping young entrepreneurs. I think about myself when I started my first company, in shared office space in New York, 19 years old, a vision, a dream, but no real meaningful experience. If I didn't have brilliant people around me who had my back, who helped coach me through that, I would have failed miserably. And so, for the last probably 10, 15 years, I've really tried to focus on giving back now, and spending my time with other young founders to try to help them succeed.

Jake Aaron Villarreal: I love that. Well, Peter, if anybody wants to find you — if you're a young founder out there, or if you have a good product view and function in an area that would be maybe interesting for Peter to meet you and share — where can they find you, and where can they find Everworker?

Peter Guagenti: Yeah, Everworker is available at everworker.ai — so very straightforward, you can Google us, you'll find us very quickly. If you're trying to find me, find me on LinkedIn — you see the spelling of my name, there's not a lot of us in the world, so you're not going to stumble across too many Peter Guagentis. There's a couple, but that's it. Or if you're a potential customer, potential hire, feel free to email me — Peter G at Everworker.

Jake Aaron Villarreal: Very cool. Well, I really want to thank you, Peter, for coming on and sharing your story today, and for the audience for listening. It's been a lot of fun, I've personally learned a lot, I know other listeners will get good value. We'll see the replies in the comments, and we'll get back to you in anything that we think could be helpful.

I'm your host, Jake Aaron Villarreal, signing off for now, but can't wait to catch up with you all on the next episode. Until then, Peter — the world, take care.

If you like what we're doing, don't forget to subscribe. Leave a review on Apple Podcasts or wherever you listen. And follow us on YouTube, where we go behind the scenes to learn what it takes to be a startup founder.

‍