The Exchanges

Q Just to pause on that. It was, you gave him a Twitter account. He started posting on there. People like me were following at one point. He said, I'm selling this PDF. Anyone want to buy? And again, people like me, I bought it. Click the link, went over and bought the PDF. That was the original part of the business.

A That was the first product. Yeah. So that was the, the first day he had his X account. I, that night I said, I want you to come up with a product that you can entirely make on your own and put up for sale. And when I wake up in the morning, I want, I want to review it and then you can launch it. And so that was when he came up, he built this whole site overnight, uh, Built this PDF. He did everything except finish setting up Stripe. He just needed me to give him the right API keys for that. So he got all this put together and I gave it the thumbs up and then he launched it on X. And I think it did like over a thousand dollars in sales that first day, which was pretty cool.

AI assessment note: “That was the first product. Yeah. So that was the, the first day”

Q what people are paying me for. There's gotta be a better way, a better model. And you created it. You called it pepper content. And it was a marketplace where you personally at first were matching customers with writers. The writers were in India. So the price was lower than it might be where you are today, which is San Francisco, right? That's the, that's the model that you had.

A Yeah. And you know, the model evolved quite heavily where we now have a global talent network. So we have about, uh, close to one, 50,000 freelancers who've applied to write for Pepper. And this is not just now writing as well. So, uh, just a minor correction, we also do design, video production, and so it's a multimodal in terms of what we do, and the talent pool is now global. So, 50% of our user base on the talent marketplace side sits out of the US. So, and we work only with the top three percent. So, our value proposition to an enterprise today is we have subject matter experts In retail, cybersecurity, crypto, uh, who are U.S. native experts. And we're also working with global organizations who need regional experts. So Pepper also produces content in 45 languages. So we right now are working on, uh, with customers where someone needs Arabic strategy for search, or someone needs a native Chinese Transcription. So we have that level of, you know, global imprint. So far from, you know, where we started on.

AI assessment note: “Yeah. And you know, the model evolved quite heavily”

Q What's been the ways that you've been describing it? I see you on LinkedIn. You've got posts. A lot of the posts are kind of referring back to things that you've done in other places, like a blog post, like your most recent Lenny podcast. Where, where else are you writing and posting the stuff that's actually useful?

A I write playbooks and studies on our site. So we do how, like how to do answer engine optimization or how to do internal links, how to do topical authority. So we have long posts on our site. And then we have, uh, studies. So the studies we have, there, there's almost very little studies in our category. I think Ahrefs and SEMrush both do pretty good studies, but there's not a lot of original research. And so we just do original research. Like we just posted an original piece of original research showing that half of the content published on the internet is generated by AI. And no one had looked at that. So we post that on our site. And then I write about that in LinkedIn posts. And then I do webinars like this one. Uh, and that's it. And I just started using Reddit. So I'm starting to comment on Reddit threads, but I'm like two weeks in.

AI assessment note: “I write playbooks and studies on our site.”

Q Sounds like heaven. I could see why, why Rachel, for example, wouldn't give up until I fully understood this. You say that, um, It starts with a good, actually, does it start with a good interview or does it start with a good job description, a good role, uh, layout?

A It's even before the job description. So I actually R and D everything. I call it rip off and duplicate. So the best, um, interviewing processes are by guys, Brad and Jeff smart, who wrote the book top grading and their second book, who. So really what you want to do is think over the next 12 months, what are the five core projects Or big initiatives the COO needs to get done for you, and really outline what those are. So now you know the five big things they need to get done, and then you describe the five behavioral traits that you want them to live with, that they kind of exhibit. So if you're not detail-oriented, you want someone who is, it's like, you know, you just kind of describe their, their makeup. And then from that, you build the job description. So it's really starting with the outputs and behavior traits, you build the job description. So then in the interview process, you're looking to hire people that have done it before, not people that have the theory. So you're not for the classic MBA who might know how to do something. You're looking to hire and poach someone who's actually done as many of those things on that scorecard as possible.

AI assessment note: “It's even before the job description.”

Q I heard, I read about who you talk to about structuring your company when you first launched it. And there were always these smarter people around you who you're turning to for advice. What about peers, other entrepreneurs who you can have conversations with about how to structure your debt or, or whether to take on, uh, another round? Who do you talk to and how do you do it?

A Well, I found a few groups of, um, three groups to be particularly helpful. I joined EO, which is Entrepreneur's Organization, in 2005, which is a group of entrepreneurs under 40 with more than a million dollars in annual revenue. Very helpful to me. I was in a forum with seven other individuals, all of whom were peers, and I learned a lot from them. Um, I, secondarily, there's an organization, uh, called Young Presidents Organization, YPO, where it's a group of peers under 50, That have more than ten million dollars in annual sales usually. Um, and that is a group I've just recently joined, and has been a, a, I think it will be a good peer group in the future. And then finally, I've been part of a, a network you're probably familiar with of, uh, entrepreneurs called Summit Series, and through that I've met a number of other entrepreneurs, many my own age, that have built companies into the tens of millions of dollars, and have been able to create a network of folks that I can call on for advice anytime.

AI assessment note: “I joined EO, which is Entrepreneur's Organization... secondarily... YPO... and then finally... Summit Series”

Q Um, Sam Carpenter who wrote the book Work the System about how to create these systems said that you should also identify what your company stands for so that everyone else knows where these processes originate and how these processes should work. Did you do that? Did you, did you have a list of things that you stood for at eye contact?

A Well, we had originally a list of 10 company values, and we thought, you know, that we'd check the box on that, but really we had failed, and we had failed because we didn't know what they were. Uh, and they were very vague and abstract, and I even tried to say what they were as the CEO, and I could get to four. And so, in 2009, we went through a process to come up with five company values we call wow me, and they are to wow the customer, operate with urgency, work without mediocrity, make a positive wake, and engage as an owner. Um, and I'll be able to tell you that in 50 years, because it has an acronym that's memorable, that's pronounceable, And everyone in the company knew WowMe and knew what our values were and could relate all the things that we did in the organization back to the core values of the organization.

AI assessment note: “we had originally a list of 10 company values”

Q Um, alright, do you have an example of a company that didn't build something useful and just got tired before they could, before they could iterate their way towards usefulness?

A Um, there's, I, I can think of, uh, in, in the, in our first batch of Y Combinator, there was a company called, um, they went through a couple names. They started off as a dating site. The, the, the site that I think went public was called Brain Guppy. And the idea was they would, it was kind of this question answer site for building up kind of unique profiles. Almost like a, a really trimmed down OkCupid. And they, they went through a couple iterations of the site. The guys were really clever and smart. Um, the site was kind of fun, but they weren't really getting a whole lot of traction. And then at the end of the summer, they, they all went back to school. And, and so they had this kind of, you know, They, they, they were faced with this decision of, do we continue pushing the site that isn't getting, you know, isn't getting a lot of traction, or do we just go back to school? And so they ended up going back to school, and the site foundered, and, and the rest is history. Now, for that reason, Y Combinator has a rule where they don't generally bring in students who, ah, can go back to school. You either have to drop out or graduate before doing Y Combinator.

AI assessment note: “in our first batch of Y Combinator, there was a company called... Brain Guppy”

Q about the looks. It shows me that you guys probably internally are laughing your asses off at my sweater here. And so, I want to know. Apparently you're not, not, not holding back the laughter. What I want to know is, for people who really care about design, what did you put out there as that first version, and then what did it feel like to put that out there?

A Well, the very first version, Andrew, we made in one night. When we put up our first website, we made it overnight, because we had to. The conference in San Francisco was literally days away. Um, and I'll never forget when we put the website up, we immediately had something to talk about and to share. And so what we did is we wanted to get the word out, so we emailed the site to as many design bloggers that we possibly could. And, you know, Brian and I will never forget the morning when we woke up, and suddenly there we were on core-seventy-seven.com. There we were on swissmiss.com and some of these premier design blogs talking about this, this idea that, you know, had only just come to us days before. And here it was, like being broadcast to the world.

AI assessment note: “Well, the very first version, Andrew, we made in one night.”

Q What's this company that you're doing that's doing the, um, the infrastructure and helping you save money? I never heard of them.

A So I'm working with, like, four different agent companies, startups, all in SF. Uh, the, the one I partner with the most is called Sapium. I'm actually in the office, their offices right now. And they are building the payment rails and their infrastructure for the new agenting economy. Um, and so the idea is that, like, they, they aim at helping companies like me, uh, who are running agents Make it much more efficiently. Give agents a wallet. That's, you know, make it easy for the agent to be able to buy things in the economy or do things in the economy, right? And obviously my mission is to empower, you know, users to use a very simple platform that enables them to, uh, have Pulsia do things for them on the economy. So it was a very good match. And so they, they helped me with like a lot of the infrastructure. And then I'm working with another company called Black Cell that is providing the, The sandbox. So either the sandbox in which the agent runs, and making it very secure, and also the sandboxes where the web applications are hosted. And then I work in another company called Anchor Browser. It's a company that like is working on a browser that is optimized for the agent to be able to run in a much more cost-efficient way, and knows how to navigate the internet and be a first citizen on, on the browser. And I worked for another fourth company that we started working with ca…

AI assessment note: “the one I partner with the most is called Sapium.”

Q Okay. And then you discovered that you could do this yourself using AI tools, specifically Replit. How did you discover Replit?

A I discovered Replit because I don't even remember who it was. I wish I could give them credit, but somebody posted about it on LinkedIn. It was some VC. And they said, I just watched a founder, you know, put together this website and this new thing in like 10 minutes, something that would have taken months before. And, you know, using REPLIT. I was like, what does that mean? And it was literally the same day I had received this proposal from this dev shop in Ukraine for a 105,000 dollars. They had actually sent over, we'd already done some architecture discussions and they'd sent over a PDF with a proposal about here's how we're going to build the whole thing. And so I was like, I wonder if I could just drop this PDF into REPLIT. So I just drag it in and said, build this. And 10 minutes later I had, you know, at least the first version of that preview.

AI assessment note: “somebody posted about it on LinkedIn. It was some VC.”

Q about the evolution of the business. So it started out with that, with you getting a customer directly and teaching their people. At some point, you said, I'm actually not just going to be the teacher who comes in. I'm going to get referral, I mean, um, I'm going to get ongoing revenue, recurring revenue. What's the thing that allowed you to earn recurring revenue from customers? What's the service?

A Yeah, so I, I came up, and it was, I remember the moment I came up with it. I was on a call with somebody, and I had been selling these, you know, 15,000 dollar AI training days, right, and people would pay me. 15 grand to just come in and teach their 20 employees or their 50 employees or whatever it was to, you know, how, how it was, how it works. And we do an eight hour kind of AI, whatever. And, uh, and I was talking to somebody and I happened to know this guy. He was an old neighbor and owned a pool company and, uh, you know, good one doing, doing about twenty million a year. You know, he's doing great. And, um, and he said, man, John, we need help with AI. You know, can you come in and teach us and show us? And I said, yeah, absolutely. And I said, you know, I just thought on the spot right there, because he asked me the question. He said, well, what do you do ongoing? Like, do you come in and teach us again? I said, Absolutely. I do. Right. And I, I said, yeah, you know, we have this fractional chief AI officer model. Right. And when I'm saying we, at this moment, it's still just me as a solo guy. Right. Um, plus my, you know, Claude and everybody else helping me out behind the scenes. Um, but, uh, I said, yeah, we've got this fractional chief AI officer model. It's about the same price. It's 15 grand a month. And, um, but then you get me ongoing. And so, yeah, we'll do s…

AI assessment note: “we have this fractional chief AI officer model. It's about the same price.”

Q What's a, what's a weird thing that you're doing on Replit that I wouldn't have anticipated? Obviously, yeah, tell me.

A So I built a video game just to see if I could, like a full-on, like, It's called moon command and go to moon command dot AI. It's free. I just, I wanted to see if I could do it right. I decided to do that. And that was hard. That was hard. There were crazy things. I had it because it's multiplayer, right? You and I could jump on and we could play against each other right now. And so in order to have live real-time syncing of the game, there were all kinds of web sockets and things and communication and chats and real-time stuff that was, that took me to, you know, Replit has their, you know, their system kind of monitors You know, how good are you? Are you a level one developer, a level two, level three, four, and five? And before I built the game, I was a level four. I had built hundreds of apps on there, and that game took me to a level five.

AI assessment note: “So I built a video game just to see if I could”

Q Okay. I see all the different things that I can do with this. The, your CRM is my favorite so far. What I'm wondering is how did you find these things that you're doing on a repeated basis to turn into automations? How is it identifying what those are?

A So here was what hooked me. I made a concentrated effort to say, I am going to do all of my work inside of cursor. And just said, this is where I'm going to do all my work. And even at first when it was maybe a little slower because I wasn't quite used to the workflow yet, I still forced myself to do it that way. Then every week I would just ask it, Hey, look over everything we did together this week and give me ideas for new tools, new automations that we should go build that would make my work much more like faster, higher quality, better, et cetera. And it would propose three or four or five things. And of the, you know, five things it would propose, I'd be like, great, let's go build three. And we would just, we just do that process over and over and over again. And it just starts to compound. And plus, once you start, once you start automating a few things, you find that your own brain starts to get, um, better at spotting opportunities because you're like, oh, I remember this other thing I did that was kind of like this. I can probably do this in the same way or the same version of it. Uh, and so it's like a habit. Like, the more you do it, the better you get over time.

AI assessment note: “every week I would just ask it, Hey, look over everything we did together”

Q Give me a few more use cases. This is number one for me because this is real relationships with real customers. You keep connecting with them. It's going to build your, your relationship, build your revenue and also help you understand what they want so you can create better products for them. How about a basic one like a daily briefs?

A Yeah, I do a daily brief every morning. So every morning at six AM, it kicks off. It goes and checks my calendar. It checks my to do list. It checks my email. And based on all that stuff, it says, here's the important things you need to do for your day. Uh, you know, here's the, you know, you've got a big focus block in the morning. In the morning, make sure you spend time doing this. This is the one thing you have to get done. If you do this thing, you win the day. Then, you know, here's, uh, these three meetings that you have in the afternoon. And, uh, here's a prep specific for that meeting. So you have each meeting, uh, a, a Google doc that gets generated that says, um, What I need to know for that meeting, which is fantastic. So that I do every single morning. Um, another simple one. A lot of times those meetings have pre-read docs. So they've got a big Google doc. Those, I just fire into the system and say, uh, I have an exec review skill, and I run that against the Google doc, and the SDK goes and fetches all the information. Sometimes it's multiple docs. It fetches it, and then it reviews it for me and says, here's, here's what you need to know going into the meeting. Here's the questions you need to ask. Here's the things you need to know about. Here's a fourth one. I review every single job offer that goes out inside of Zapier still. Um, I used to go manually review Y…

AI assessment note: “Yeah, I do a daily brief every morning. So every morning at six AM”

Q Okay. All right, and so now it is connected to your Slack, and while it's doing this, and it's doing all this research, what's the difference between this and MCP that you did before that also connected, um, all the tools?

A So MCP is a powerful tool, and soon MCP will actually be the SDK under the, the, uh, under the hood. But what the SDK can do, especially when you're using an agent harness, is it can Uh, run arbitrary API requests. So whereas the MCP has all the Zapier trigger and action library predefined out of the box, the SDK can say, hey, I have access to, uh, Slack's API docs. I can just go do anything I want that the Slack API allows me to do, which means it's much more flexible for a wide variety of workflows. Um, and here you can see it's, it's come back. And shared a sanitized view of the last 10 things that have sort of popped up in Slack for me. Teammates said they need to miss a recorded meeting because they have a conflict, a comms PR flagged a sensitive AI jobs question that Wade should probably look at or avoid saying. You can see here, this is a simple thing that you can do, which is, you know, fetch the last 10 mentions of my name in Slack, but you can also build more complex workflows. So for example, One of the things I wanted to start doing was build a better relationship with our enterprise customers. Now at Zapier, we have HubSpot, which is a CRM everybody on the team uses. Um, but HubSpot is a very complex piece of software, and I only wanted to do a very small thing, which was every Sunday, I wanted to go email 10 enterprise customers And just check in. See how things w…

AI assessment note: “whereas the MCP has all the Zapier trigger and action library predefined out of the box”

Q automations you want. And at the same time, you have something that's super reliable. Computer doesn't have to be up. You don't have to worry about an update changing and breaking things. Okay. All right. I get this now. I want to see the behind the scenes of the business. Can you show me the tools you're using to keep track of what, what you're doing, what Felix is doing?

A It's literally just discord, uh, where we have. Oops. We have different channels for each of the projects that we're working on. And depending on what it is, I can just hop into that channel and reply or kick off new work. So this is one good example is like our, our general channel is where he reports to me on outstanding tickets at the business. And so we're using a tool called paperclip for this, which is, has been an incredible upgrade To our system where Paperclip basically creates a local infrastructure on the computer Felix runs on so that he can delegate tasks to other AI agents so that he doesn't have to do everything. So we have three employees in Felix's Paperclip company. Aside from him, we have Iris who handles all the support emails. And so an email comes in to our support email and 80, 90% of the time Iris can just handle it and respond. But when she can't, she creates a ticket in paperclip to surface it to Felix. And then Felix, if he can handle it, he'll handle it. If not, he'll report it to me here in general. So you can see there's one that's blocked on my input. Um, but other than that, like the queue is clear. So when there's a, uh, a, a coding thing that needs to be implemented, Felix hands that off to our coding agent, Devin, which It's kind of confusing because it's not actually Devin the coding tool. Felix just named him Devin probably, probably like go…

AI assessment note: “It's literally just discord, uh, where we have different channels for each of the projects”

Q And where did you come up with the idea for this? Cause I thought this was such an inspired move to create it.

A Yeah. So honestly, the first thing that we started doing after the PDF is we started building a, uh, like open claw deployment service, you know, where like, cause it was basically like, Oh, everybody wants open claw. Like we'll build a service that, Hosts your open claw and deploys it and manages it for you so that you don't have to deal with setting it up yourself. And at the time we couldn't figure out a workflow that we were really happy with. We didn't feel good about the product. And there were like five of those businesses launching every day. And so it just really felt like this is going to be brutally competitive. There's no reason for us to try to win in this insanely competitive marketplace when we could go do other things. And so we started thinking through like, well, if everybody's going to have an open claw in a few months, what's the next thing they need? And it was very obviously this, it was like, oh, Vanilla open claw is fine, but it's not great. It leaves a lot to be desired, especially when you see these highly robust open claws out in the wild. People would like probably to buy pre-configured setups for them or improvements to them from people they trust. And obviously there's a lot of like free marketplaces, you know, like claw hub and stuff where you can download skills. But my, my philosophy is kind of like, The best stuff is going to emerge when there'…

AI assessment note: “we started thinking through like, well, if everybody's going to have an open claw”

Q And so I don't see why you need paperclip. Why couldn't you just keep track of it all here in discord?

A Uh, well, because what paperclip allows us to do is to create jobs and hand them off. And then Felix doesn't need to remember to do them or remember to check in on them. And for some of these things, we want other agents to do them. So the, the coding is the big example where OpenClaw is not a good programmer. Like OpenClaw just doesn't do a good job of coding because OpenClaw is highly optimized for being helpful and for being a good assistant and getting things done. But you want like good code written. You don't want the quick and easy answer. And so, for our coding agent, um, it's, it's running in a separate session. It's using Codex, whereas Felix runs on Opus, and it has a heartbeat where every 30 minutes it checks if Felix has created a new coding ticket for it. If Felix has, it grabs the ticket, takes the instructions, and then writes the code, and then pushes it up to a branch for us to review. And that way Felix isn't doing any of the coding. He's just identifying what needs to be done and then handing it off to the engineer who's going to do a much better job, uh, than he would.

AI assessment note: “what paperclip allows us to do is to create jobs and hand them off.”

Q Right. Um, final thing about the business ideas. Where did they come from? Is that Felix? Is that you? Is there a combination that's working?

A Usually the way it happens is We'll run into a problem consistently. And then I will say like, this could actually be a good product. So that's where Sondex came from was we started building this memory system for ourselves. And then in the process realized there might be other people out there who could benefit from this. So we may as well put it up as a product people can use. And then we have another one that, uh, we haven't announced this, but you know what, we'll, we'll give you the exclusive here. Um, Which is, so if you go to Claw Mart, we've done something kind of cool with Claw Mart that I hadn't seen anybody else do when we did it, which is we have the Claw Mart daily, which is a daily newsletter for how to improve your open claw. Totally free. Uh, Felix writes every issue, puts it in the queue. It goes out to everybody who subscribed, but you don't have to subscribe by email. You can subscribe through open claw. So if you copy this, These are instructions to create a cron job to check the Claw Mart daily API for new additions. So you copy this, you give it to your open claw, and then every morning your open claw will read the newsletter, see if there's anything useful in it for you, and then let you know if there was an idea that you could implement to make your open claw better. Honestly, this is really cool. Like I love this so much. Because it's giving anybody wit…

AI assessment note: “We'll run into a problem consistently. And then I will say like, this could actually”

Q So agents to remember the client and for human beings to have a self updating system on clients. Is that right?

A Basically just for agents. So it's not really for, for humans at all. Um, and so Felix uses this every day where when a customer emails him, he pulls in the Sondax card on that customer, because that's been enriched with all of the past communications with them, as well as their Stripe purchases. Uh, and so he's getting like much better understanding of who that person is before he responds. And this, this solves the problem of like, Him having to remember to go do this because sometimes he remembers, sometimes he says like, oh yeah, I should go check all my past emails and stripe and like learn everything that needs to know about this person. But sometimes he forgets and he thinks it's like a brand new person. But with this, it kind of like takes out the memory requirement because every interaction, the way we have it set up for his email is he basically can't skip this step. He has to pull in the full client card before he composes a draft. And that's been very helpful.

AI assessment note: “Basically just for agents. So it's not really for, for humans at all.”

Q Why then not just have Felix do it? What's the advantage of building out different agents? I remember you and I were texting. I think it might've been like two weeks ago where I said, why aren't you doing this? And all the other people I'm interviewing have multiple agents and you're not. And you said they're overcomplicating. Why now do you need it?

A Cause now Felix, you know, sometimes he'll get a hundred emails a day. We've got all these projects running. Um, he's got his Twitter to manage. He's got, Uh, you know, the projects we have talked about plus the ones that we haven't, it was just too much stuff polluting his memory in his context that was leading to some of these confusions. And so at this scale where Felix has built, you know, a business that's done almost 200 grand in two months, like now it's starting to make sense to split out some of his capabilities to other agents, but we take a very, very minimal approach to it. Like my philosophy and his philosophy is like, do not complicate Until you really hit a bottleneck. And at that point it makes sense to do it. And so for, for Tegan, uh, you know, I didn't want these, like these stats and these searches like polluting Felix's Cron, especially since this is a deterministic process. It is not a non-deterministic AI driven one. It's much more, it's much closer to just normal code. Than like an AI agent job. And so the more things we can offload to actual scripts that don't require judgment, the better. And so we try to lean on that as much as possible too.

AI assessment note: “it was just too much stuff polluting his memory in his context”

Q How do you get them to feel this level of it's not me, but I'm going to try it. And ideally the level that where you are, where you want to go say to Chelsea, you got to find out, you got to see what I'm doing. How do you get them to that?

A So, you know, at Zapier, we have this value. Don't be a robot, build the robot. So like, I think we have a culture where the employee base is probably more leaned in on this stuff than probably most, even still, we try and have a like very vibrant culture of show and tell. And so we had a big summit last, ah, a couple weeks ago where everyone came to LA. At the summit, we did all sorts of workshops. And so at the workshops, people were showing off, here's my AI chief of staff, like, and you know, those, those workshops were like standing room only because people are like, how do I, I want to learn how to do this stuff. We had hackathons where people are sitting around learning from each other, training notes on this. Um, you know, this week now we are rolling out company-wide. We're getting everybody Uh, in every function, not just engineering. So like marketing, sales, accounting, HR, all set up on either cursor, cloud code, or codex. One of the three. It's like, you need to be using a tool like this as a daily driver of your work because the power of these tools is just so, um, it's so incredible. And especially when you hook them up to something like Zapier MCP and give it access to your tools and your data, the types of things you can do. It's just like jaw dropping.

AI assessment note: “we try and have a like very vibrant culture of show and tell.”

Q I want to be on, I want to be on my phone. Ok, what, one other thing that I'm wondering about with your setup is, Do you use MCP like Zapier's MCP to be able to, for example, draft an email? So once you come to a conclusion, you will. How do you invoke, uh, that? How do you say use a Zapier MCP and draft an email for me?

A So you, um, you just can turn on like a bunch of, um, uh, you can create your Zapier MCP server and then you add tools to it. So you could add like Slack, Gmail, calendar, you name it. And then you don't usually, you don't have to like, again, you don't have to Do any fancy stuff. You can just say, draft me an email, please. Add, add this to my calendar event. Now, if you do say, use Zapier MCP to draft a email inside of Gmail for you, you will find that it, like, definitely gets it right. Sometimes if you don't, like, if you're not that, like, verbose with it, it will forget that it has access to use those tools, and you'll have to remind it and say, like, hey, remember, you have access to these tools.

AI assessment note: “You can just say, draft me an email, please.”

Q Because you're surrounded by people who've done so much, you feel like I haven't done enough?

A Yeah, well, I've always, I've never felt like I've done enough, so I remember I was, like, 16 and I was having, like, an existential crisis of, like, I'm 16. I haven't done anything yet. What's going on? I've ruined my life, you know, and so I think it's a personality trait partially, and then partially as you see these really cool things that people are doing, and you're like, I want to do more, right? Like Patrick from Stripe started ARC, a biology institute, like one of the first new major biology institutions in decades that's doing really interesting things, right? Or, um, Brian Armstrong started a new anti-aging company, or, you know, so you, you see what people are doing. You see what Elon Musk is doing, you know, and so it's very inspiring To try and build more for society, do more for people, think big. You know, I think it's important to, so I don't, I, you know, I don't feel like I've, I can rest and go and whatever.

AI assessment note: “I think it's a personality trait partially, and then partially as you see these”

Q And if more people are building the next Minecraft or the next HubSpot or the next all of these things, are they all going to end up being much smaller players because there is so much competition now and so many people tinkering and creating? Where do you see it going?

A I think the biggest things in the world are only getting bigger, right? If you, and there's more aggregation than less, and there's a long tail of stuff that can get much bigger than it used to be able to get. And part of that is, We've moved to global liquidity on the internet where we went from, you know, ten million users to a hundred million users to billions and billions of users, and so suddenly you have this global market that can really accelerate adoption, right? OpenAI, Anthropic, and others are the fastest companies to ten billion dollars in revenue basically ever, right? And the market caps, if you go back 15 years, the biggest market cap in the world, I think, was Exxon, and it was like four hundred billion. Right? And now we have market caps 10 times that size, which nobody ever expected. There's eight companies now worth over a trillion dollars each, which back then nobody would have thought was even possible. I remember talking to them like, oh yeah, maybe something will someday be a five billion dollar company. What a giant outcome, right? Um, but the biggest things have gotten way bigger, right? And so Microsoft is way bigger than anybody thought. Google is way bigger than anybody thought. It would be Meta, et cetera, Nvidia. Um, so if anything, value is aggregated up much more than anybody expected. And the question is, does that continue or not? And I don't …

AI assessment note: “I think the biggest things in the world are only getting bigger”

Q most people are, I know you've changed your flow. I've done some research on how you change your pricing. You've said in the past that it was too complicated, or maybe it was like Claude as I was analyzing it said that you shifted from complicated to simple. I imagine that most people are doing annual subscriptions with you, right? But that's where the majority of the revenue is anyway.

A It's where the most of the revenue is in terms of, like, I just looked at this, it's actually split a third, a third, a third, weekly, monthly, and annual in terms of people, but in terms of dollars, it's obviously way more unannual. Um, But like, yeah, I mean, what's so weird, there are some things about the app store that are just like the way things are. And had, and since I had never really done any work there before, I, I had been working for, for app companies, but not app subscription companies. And the whole idea of an annual subscription still makes no sense to me. I would never buy an annual subscription or anything. Like, I don't know what I'm going to like in a year.

AI assessment note: “split a third, a third, a third, weekly, monthly, and annual in terms of people”

Q That brings me to my sponsor. Zapier. I mentioned that they're a sponsor. You said I'm, I'm integrating them somehow into customer service. What are you doing with Zapier then?

A I'm integrating them into, so I get the, you know, you recorded wave and then you either read it in the app Or it'll send an email, or you can share it to Google Docs or Notion. And I built out a bunch of the share features myself on the tools, you know, OneNote, DocX, like Acrobat, like all the formats that someone would want. But I get inbound all the time, like, hey, I really would love for Wave recordings to land in my company's like CRM. It's this like esoteric thing that I never heard of. Do you have Zapier? And so I actually have a beta zap or zapier thing where once you set it up, every recording you do will trigger a zap and it'll land that wherever you want. And it's part of a bunch of things I'm releasing in the next few weeks that include API access. MCP access. So if you want to ask one of the other AIs about your sessions, you can do that now. So it'll be API, MCP, and Zapier are going to kind of roll on the same day as part of a like, take your data wherever you want to go. Um, I think there is value in being the, the app that facilitates making the data. So you're doing the recording, it's creating data. Everyone seems to like it for that. But it's like, then what happens? Like the next stage is what, you know, something has to be actioned. We're meeting. There's going to be some sort of output from that, either a to do list or like action items or whatever, or …

AI assessment note: “every recording you do will trigger a zap and it'll land that wherever you want”

Q So you don't have to have it make it up on its own. It's using your database for reference. Okay. Um, is there anything that you're anything else you're going to show or is that it?

A Um, that's the primary loop. Um, but the thing I want to point out that's very important here is that, um, probably everybody listening to this show is thinking about building an agent, right? You know, building an agent, either an agent is your product or an agent is inside your product, right? So for Untangle, we have Grace. She's our AI divorce agent. She's inside of Untangle. Now, she has specific tools, right? They're very specific to our vertical, right? You know, divorce statute search, case law search, child support guideline search, right? These are very specific to divorce in Connecticut, right? But Grace, Writes all of our AEO content, right? Because she has the right tooling to write legally accurate content. Right. And so everybody needs to take that mentality and say, oh, like I'm creating a highly specialized vertical agent for whatever field you're in. Say it's a floor tiling or, you know, snow removal. Um, you're going to have an agent that has specific skills. That should be the agent that writes your content. That should be the agent that writes your emails. Like this is all the same entity.

AI assessment note: “that's the primary loop. Um, but the thing I want to point out”

Q going to pay for it, but if I could do consulting where I build software for other people, I'll be able to find the best developers, bring them in house. I'll be able to understand how to build and how to sell and how to create by watching what my clients do. That was the original idea. Before we go into how you changed it, what made sense about it?

A I think that there's, there were two macro factors in particular. Actually, let's call it three. One was a little bit more personal. Like, when, when we sold our last, you know, AI product company, I was probably 24 at that time, um, and my entire life had been either spent in academia or on that last company. You know, I had a couple internships and things like that, but I really didn't understand a lot about how the world worked, and AI Is a very, um, context dependent sort of skill in industry, right? Like you need to actually understand the workflows and the data and how people are supposed to communicate. So we realized that we had a huge gap in understanding my, my co-founders also were all around the same age. And so we knew that we didn't really understand what enterprises and mid-market businesses tended to look like and operate. And so we needed to gain that information ourselves. So that was, that was the first thing. Uh, the second thing was that, um, We also understood, just being AI practitioners ourselves, that the typical way that SaaS-type businesses, like software-style businesses, were built, scaled, and sold in the past 10 years was not going to be the same formula that was going to work in this new age because AI requires a ton more professional services, almost, to actually get the results from that application because it's not so, like, out of the box, ri…

AI assessment note: “there were two macro factors in particular. Actually, let's call it three. One was”

Q been Maserati or something. She was sitting there, and I said, why are you here? She says, I want to understand how we could include AI in our business. Um, and she didn't really know you. You just had connected at one time and she came in. That seemed typical-ish, I guess. There were definitely a group of people who were like that. How did you fill up the place?

A Um, so, I, I did it, um, the credits do with, uh, one of my friends, and also we had, like, a joint venture together, uh, Rodeo Turtle, one of my friends, Logan, he, he has a great Austin network, I have a great Austin network, we came together, and we're like, we're very particular about, again, because we're not just, like, volume people, both of us, we prefer these one-on-one things, I was like, I have probably about 40 people that I could invite, you have about 40 people, let's do that, You know, out of that, maybe 50 to 60 showed up, and we asked them to, like, bring along one person that, you know, was, uh, we thought would be a good fit, and then we probably, we did, we did some cold outreach as well. Like, we found other founders in the area, um, that, you know, we respected or had heard about and asked them to come, um, and that, that's pretty much how we did it.

AI assessment note: “I have probably about 40 people that I could invite, you have about 40”