The Exchanges

Q in March, uh, 25. A number of people have made a point of saying this year is a year of, um, uh, AI agents. Love your take. I mean, do you, do you feel at this point in the life cycle it's, it's more hype or, or do you get the sense that there's some genuine green shoots and you're, you're really optimistic or bullish, uh, on, on the segment?

A I mean, I think, unfortunately, I think the, the phrase agentic has become overused, and it really is more hype than reality, but I do think there, the core of the, the shift to agentic is, is really enterprises are looking for demonstrable ROI. And the challenge of the sort of co-pilot model, there's been examples of where, um, a private equity firm has deployed a co-pilot to 20,000 employees across their portfolio, saving on average, 48 minutes a day. It's a fantastic end outcome. However, it's not 48 contiguous minutes. So it's not like I got an extra hour to my day. I had an extra two minutes here. I got an extra 15 minutes there. So I, I'm not fully newly productive with that. So it becomes hard to measure the ROI When you have that kind of squishy return and squishy value add, um, with the Gentic, I think the, the real core is that enterprises are pushing for that agent. I can measure exactly what the outcome is. I can measure what it took to start the process. I can measure what, uh, what the net improvement was on completing the process. I can, I can measure the, the improvement in throughput and all of those measurable. So I think Really underpinning this whole movement is that notion of generating measurable ROI as much as it is about having this autonomous agent that goes off and does whatever you'd like it to.

AI assessment note: “the phrase agentic has become overused, and it really is more hype than reality”

Q run into a lot of friction because you're talking about improving people's lives versus displacing them. Um, Bunch more questions on Sancero. Before we do, we'd love to learn a little bit more about your background. Uh, I know you've been an entrepreneur. I know you, you've, you've, you are, have been an angel investor. How, um, what was your, your career trajectory leading up, uh, leading up to Estes?

A Sure. So the, the way I generally describe it is that I'm a technologist, entrepreneur, and investor in that order. Uh, I started coding at 14 years old on a TI-eighty-five graphing calculator. I started my first business when I was 19 years old, my second at 22, um, after having gone to work for a dot-com that failed during the dot-com bubble. Um, and started my third business, uh, that I sold about 10 years ago. Started that in, in 2005. Um, I, I dabbled in angel investing, um, in the, called, 2010, 20 11 time frame. At that time, I was running two businesses, um, which if you, if you're going to be an investor and do it properly in, Private markets and especially early stage startups. It's not the best idea to do it when you have zero free time. Um, it takes a lot of time, a lot of work. It's, it's really being an angel investor is, is just as much work as being a venture investor. Um, so I, I dabbled in it in the, in the 2010, but once I sold easy seat in, uh, in 2015. Uh, we sold to a strategic, so I didn't have a long earn out. I didn't have a lot of involvement with the, with the company. Um, I got, got involved in angel investing again. And what I realized is exactly what I just described is that it's a lot of work and it can be a full-time job. If you're going to generate quality deal flow, if you're going to find quality deals and you're going to do adequate due dilig…

AI assessment note: “the way I generally describe it is that I'm a technologist, entrepreneur, and investor”

Q don't know, oversimplify and put AI into two categories, at least, uh, for now, and I'll call one of them co-pilots, where they're assisting the user be more productive, and I'll put agents in the other category, where, where you have, um, a focus on automation. Or, is, uh, Centero focused on, on either one of those two use cases, or do you find your, your, your, your pretty opportunistic?

A So we're opportunistic. We'll look at both. We like both use cases. I think one of the challenges with agentic, um, is simply the it's, it's hard to build a complete solution, right? And within the, in the co-pilot use case, you've always got sort of the human in the loop or you're, you're making their job easier. You're making them more efficient. You make them faster, but they're the ultimate arbiter of success or failure in the agentic use case. It requires a little bit more caution or in, in a deeper solution because you're taking the human out of the loop. You're expecting the agent to take automated action and deliver automated results. So what ends up happening on the agentic side is the, we tend to, we tend to require companies to have a much deeper solution than you would in the co-pilot use case.

AI assessment note: “So we're opportunistic. We'll look at both. We like both use cases.”

Q Fantastic. Um, so if we can start unpacking your experience as an entrepreneur, um, sounds like you're in this great position as you have a combination of Uh, you know, not a great outcome. Um, how's that, how's that informed, um, you know, how you evaluate, uh, businesses today?

A Sure. So one of the things that we look at when it comes to evaluating business is really, it's a, it's a one foot in front of the other process, right? It it's, you, you go from zero to a million, a million to 5,000,005 million to ten million dollars, but you don't sort of skip steps in the middle. And there's this common misconception in venture capital that you somehow just wake up one day and you're a hundred million dollars a year company. While that does happen, you do have your, your edge cases like open AI of the thousands of companies that are successfully venture funded every year, they follow that one foot in front of the other process. And that's no different than being a bootstrapped entrepreneur and building any company from the ground up is that every day you work, you wake up and you work on Taking that next step and moving the company forward. Really the only difference with venture capital is that maybe instead of walking those steps and, you know, sort of crawl, walk, run, we, we can accelerate the rate of progression of crawl, walk, run, and maybe we can run a little bit faster with, with venture capital and venture backing. So, um, but we still fundamentally believe your goal as a venture backed business is to build a great company that stands on its own using capital to accelerate that process.

AI assessment note: “one of the things that we look at when it comes to evaluating business”

Q Um, you know, so said another way, it sounds like we're, we're in early innings, uh, but there's promise. Um, you know, in, in the meantime, I heard you reference e-commerce, you know, are there any, uh, industries or, uh, uh, sub industries that in At the moment?

A So we, we tend to, we tend to like the, we call it the jobs not being done. You know, it's the, it's, it's places where they've historically been under-resourced and under-invested. So think about things like compliance and human resources. A lot of you are sort of cost centers within an organization that The work isn't getting fully done anyway, right? Or it's not, you're not fully dedicating the resources that you need because it's not cost effective. It doesn't generate more sales. It doesn't lower costs of your organization, but it's important work that needs to get done. We're excited about those types of opportunities because it's work that needs to get done. That's not being done. It's not going to take anyone's job away. It's actually a place where companies will realize the benefit of being better in their HR processes, being better in their compliance processes, uh, being better in their recruitment processes. You'll realize the benefit of it, but you're not going to have that sort of trade off of, you know, are we letting people go or are we reinventing the wheel? Like it's something we can hand off for automation pretty easily.

AI assessment note: “So think about things like compliance and human resources.”

Q Okay. You mentioned, and so with that, you mentioned, um, product market fit, uh, arguably the holy grail, uh, for early stage company to achieve, um, you know, as, as you reflect back on your experience, um, your time with Santero, you know, is there a profile or characteristics in the evolution of a team and its product to get closer to, uh, product market fit?

A I mean, I think the, to steal the quote from Jason Lemkin today, um, is the, you know, you've got product market fit when you don't know where the customers are coming from. You know, when you're, when you're early in the life cycle, when you're early in the life cycle of the business, you know, exactly how you got the lead, you know, exactly the process you took them through. You remember all the conversations, all of the negotiation. And when you've got product market fit, it just kind of happens, right? Like you've got process and structure and, and what have you. And we have a similar thing at Centero where, um, I couldn't specifically tell you how we generate all of our inbound leads for Uh, for companies because it's, it's the culmination of a long process of, uh, of kind of developing product, developing services, developing messaging and go to market strategy. And the same thing happens with companies is that, you know, you sort of know product market fit when you have it, but I love that measure of, you know, when you, when you look at the customers that came in and you're just like, they just appeared, they, they found us or, you know, um, they got referred to us.

AI assessment note: “you've got product market fit when you don't know where the customers are coming”