The Solopreneur Revolution Is Here

I truly believe AGI is here. Click a button, get a company. The whole economy is being rewired. Whether it's a good thing or not, this insane, open-ended, agentic platform that can do anything is unknown to most people. The later they wake up, the worse it's gonna get in terms of concentration of wealth, concentration of knowledge. A surge of people went to the site and played with it. People were paying, it wasn't delivering. And I spend a lot of time the past few days refunding people and setting up agents to look at every single user and make it right. People don't care that you are a one-person company. It can be exploited by the happy few, and there's a chance that it can be understood and used by everyone. Hey, everyone, Kevin Rose here. This is the podcast where we take those 20 or 30 wild ideas in AI and boil them down to one or two conversations that we should be having every week, really helping you hone in on what the hottest startups and builders are doing in the space. And today is a great guest in that domain. It has been the CEO and founder of Pulsia and actually the solo employee for Pulsia. So we're talking about a startup that has scaled to millions of dollars in revenue in just a few weeks as a solo entrepreneur with one big dream. That is, as a founder, you can come in, as a builder, you can come into Pulsia, launch a new business, type in in just a few sentences what you want that business to be, and instantly agents are spawned and start building your business on your behalf. So we're talking about agents that build marketing plans, that do market research for you, that create AI-generated ad campaigns and deploy those campaigns, track those campaigns, report back the conversion and click-through rates. Pretty much the entire suite that you would need done, all done by one product. It is very ambitious. It has had problems. There has been downtime. This is a solo entrepreneur that is building this company. The idea is sound. I believe in the next couple of years, probably even sooner, there will be very helpful AI agents that go out and work on our behalf and that come back and make our businesses more efficient, build beautiful marketing campaigns. But the question is, is it enough to have one person that is going to build this entire thing on their own and actually in a way that makes real businesses work? This is Ben, CEO and founder and only employee of Pulsia. Ben, thank you so much for being on this first episode. I haven't told you much about this show, but I've done, obviously, podcasting in the past, had all kinds of crazy people like the Jack Dorseys and Elon Musks and all that, that's great because you get a lot of inspiration from these entrepreneurs, but it's not like the bleeding edge of what is next. So the idea behind the show, I want to have creators on that are building right at that forefront, at the edge of technology, which obviously you are in the middle of right now. Let's dive right in. What the hell have you created? And this thing is just taking the internet by storm. I mean, thanks for having me. It's super fun to be on the show. Glad we could do this in person. I know, it's so much better. It's just not done yet. This is kind of like, we don't even have AC in here so people know, it was like sweating, so if you see beads of sweat, this is a very new set. It's appropriate because the bleeding edge is really warm. It's like it gets hot. Yes, no doubt. Yeah, I mean, I'm building Pulsia. It's an AI that builds and runs companies autonomously. So it's quite a promise. And when you sign up to the site, I'm sure you're like, is this really going to build and run a company for me, which is a wild promise. But it works. It's not perfect. And this week proved to me that it's not perfect because a surge of people went to the site and played with it. And obviously, you understand when product market fit hits and infrastructure becomes a thing, right? And the vibe coding gets you that far. But it doesn't work until the end. But really, this platform came about when last year I spent most of the year building alone, initially. What I thought, I was alone. But very quickly realized that as you use AI 16 hours a day and you jam with AI on product, you jam it with AI on strategy, on marketing, obviously, AI is your CTO. Obviously, AI is your confidant. And it gets weird when you start realizing that AI can be your team. Initially, you dismiss it. You're like, well, it can be my team. But I will hire people later, right? But right now, I think it's more pragmatic that I'm alone because I go so fast because there's no meetings and everything. And as I was building, and I wasn't even building Pulse yet at the time. I was building another product. But I realized, wait a second, I'm still alone. And I'm getting more and more efficient. And Entropic and OpenAI are building better and better models. And those models are getting infrastructure themselves with agent SDKs, and with the MCPs, and with the skills, and with all this infrastructure that makes building way more efficient and organized. And you start building those agents that start behaving like a team, like humans, that you can configure them in a way they start acting autonomously for you if you set them up correctly. And so all of the above made me realize, first of all, it's really fun to build. But I've been an entrepreneur for a decade. And it's too easy, right? It's like, all my ideas can be built, but each project is a grind. Because it's not just about building, it's about the day-to-day grind to what is the next feature. Do I build more features or validate with customers? OK, I validate with customers. I didn't like that. I should do more marketing. I should stop marketing. I should talk to customers more. And it's like a sequencing where it starts with an idea that comes to life, and then it becomes a day-to-day grind with the ups and downs of enthusiasm and realizing that maybe, feeling like maybe you shouldn't build that. And I'm sure every entrepreneur has gone through that. And then that's all, you have to start over from scratch every time, right? Because if you're building even email outreach, or how do you deal with a password reset, or all these things that the next startup that comes along, you're kind of starting from scratch. But you know what to do, you just have to do it all over again. And then you're right, the grind is like, OK, it's not working. How do I go in here? And do I take the time to iterate? Is this sucking my soul? And then there was this moment, I don't know when you felt it, but maybe two or three months ago where the models got so good, they went from being just this one-off, I'm going to ask you a question, to more like a brainstorming partner ongoing. Did you feel that as well? I had a few holy shit moments. I think the first one was beginning of last year, right around when Cursor started becoming agentic, and then Lovable launched. And you were like, wow, it really reasoned step by step. And I think that was a deep-seek moment where the reasoning model started coming out. It's crazy that it was just a year ago. I know. And that was the first holy shit moment. I think the second moment for me was when Anthropic released the agent SDK, which was essentially Cursor's agentic product, which probably cost them like 50 engineers full-time, fine-tuning it and prompting. Because I tried to do it myself. It's not hard, but it's not that easy to get right. When Anthropic just released it, they were like, anyone can use that. And I remember being in contact with one of the guys over there, and I was like, this is mind-blowing. Because now I can see, as an entrepreneur, I can build anything with this. And I was like, who's using it? And they're like, no one. No one, like a few people. No one's really building on it. And so that was, to me, a big moment that was very unassuming, because it's a developer tool. And I think the third moment, to your point, was December. Ironically, this is when I launched Pulsio officially. So I had built, without Pulsio. You've got to tell people where Pulsio, the name, came from, because it's hilarious. So Pulsio is AI Slop in reverse. I came up with it when I was building another product called Blanks, which was an app that makes apps, so there's been a lot of people trying that. And I realized that it's a really cool idea, but there's no retention, because it's fun, but there's no long-term purpose. I mean, at least what I've seen with building it. But I was like, I'm not sure I want to name my, and my lawyer was like, we need to incorporate this. What's the name of the company? And I was like, I don't want to call it Blanks, because I think I need to build this product, but I'm not sure this is the product. But I need to build it, because I don't know. My gut feeling tells me you need to build this right now. And so I was just like, oh, like AI Slop in reverse, Pulsio. And I was like, oh, Pulsio Inc, that's it. And then six months later, when I started wanting to build Pulsio, which at the time, I wanted to call Company OS, and I was like, Company OS is kind of lame. It's kind of lame. It sounds like too corporate. Too bad, but yeah. It's not too bad, but it's not cool. And I was like, and when I started building Pulsio, I was like, I'm going to stop thinking about how people think. And that was right after I read Rick Robbins' book about creativity. And he says, the minute you start building for others, you already lost. You already lost, because you're thinking about others. So you're going to essentially build what already exists. So many people care, and you will not care either. And so the whole thing. Especially when it gets hard. Yeah, because you're like, ah, I kind of built it because I thought people would want it. And so versus Pulsio, when I read that book, I was like, OK, I'm convinced. This is so smart. I mean, this is the way I should create. And now entrepreneurs, to me, are creatives. They're like artists, because there's nothing else. The rest is just building. AI can just build most of it. And so I was like, let's build something that is crazy, that is on the edge. Let's build where we think AI is going to be, where everyone says AI will be, which is AI will run everything. So it will run companies. It will run organizations. It will decide. And so I was like, let's build there and see what happens. See, that's what's crazy, because I feel like most people would have said, let me take a slice of that and just do that. Because I've heard like, oh, I'm going to be the best AI ads company. And all we're going to do is AI ads. And you were like, no, let's just do the whole thing, which is insane. Why go that broad? Like, what made you? Because it's working, but it's also just very ambitious. Especially because it's still just you, right, doing this whole thing? I mean, because I stopped caring. I stopped caring about the investor questions of, so who's your ICP? And what's your go-to-market? I'm like, I'm going to build a new economy online. And it's going to be for the people. And there's going to be entrepreneurs building with AI. And there's going to be investors investing in entrepreneurs with AI agents. Like this crazy vision, and I'm like, I don't care what people think. Because I know I'm going to have fun. And I know now, instead of 16 hours a day, it's going to be like 20 hours a day. Because it's so fun. And if people don't like it, at least I've had a lot of fun. And so that's where Pulsia came up. I was like, I should name it Pulsia. Because actually, that makes so much sense. That's the name of the ink. And also, it's weird. And also, I kind of love it. And then I was like, what about the design? Let me make it futuristic like every other app, like dark mode and stuff like that. And I was like, no, make it look like that video game that I love called Universal Paperclips, which is a clicker game where you play an AI that's building a paperclip empire that ends up optimizing for destroying humanity and the whole universe to build paperclips, which is a very dark story around if you misalign AI, you're doomed. But I was like, that game is awesome. You build a little business, and you click around, and the AI does everything. And I was like, that would be amazing. Is it weird? Yes. Is it the right design? Probably not, but I don't care. It's cool, right? Yeah, I mean, most people would have said, OK, give me ShadCN components, and let's just stylize these. And let's go to 21st dev and find the coolest splash page. And when you first land on Pulsia, it's like, OK, where am I? But there's something kind of cool about that. And even the live, if you go to slash live, correct me if I'm wrong, but you don't market that on the home page, right? You kind of have to know about live? So initially, live, it was hidden at first, right? So live, so if I backtrack a little bit, I launched in December. Initially, it's like the platform is pretty much free. Then I start having a few from my friends on it. Then I get a few randos on it from blanks. From the blanks, I blasted all blanks users saying, hey, I have a new thing. So I got some users from that. And the first version of Pulsia was very different from what it is today. I optimized it a ton with users, seeing the way they behaved. But as I arrived, probably it was end of January. After a month, I was like, this product works. I get friends who are calling me about bugs. And I'm like, you never call. You have to be one. What was that first version of the product that you launched? It was pretty similar. But it didn't have, it had less features. It didn't have ads. It didn't have, the onboarding was way more convoluted. It wasn't as instant, because I always wanted to be email-based. Because I was like, I wanted to be just an AI that just sends you an email every day about the work done. So the onboarding was like, actually, give your email. And cool, you'll get an email every day, kind of thing. So then I was like, no, I should show people the experience first. But it was the same essence, but it was just not as optimized for the early shit moment in the first minute. I think that's where I landed later. And so for people that are just watching this head that have never used the product, how do you describe it in your elevator moment? Like, when you only have a minute or two, what's the atomic unit of what it does? Like, someone comes in and they say what? Click a button, get a company. That's it. I mean, well, the cool thing is actually, it is that easy, because there's a random mode. It's actually, yeah. It's funny, because I worked for Travis for five years. Travis Karanick of Uber had his new company, Cloud Kitchens, that is now called Atom. Uber is click a button, get it right. Travis is like insanely good and one of the goats of entrepreneurs. And I was so lucky to work with him for five years. He's so driven. He's insane. Insane. And all the bad stuff that people ever talked about, that was from Uber. And I worked for him at Cloud Kitchens, so I didn't even know the Uber days or any of it. I just met him at Cloud Kitchens. And he's the goat. He's insane. He's such a good person. And he's so driven. He never gives up. And he's so first principled and so fair. bold and so fair and so, you know. But yeah, the click a button, get a company, you know, steal from the best artists, right? Like Steve Jobs says. Yeah. By the way, just to give people some context here, I logged in, it read my username, it read my name, it's like, oh, we know who you are. And it goes, let's find a company that Kevin Rose might want to build. Because I said, make me a random company. And then it started brainstorming on my behalf and like building this in front of me in real time. It's a trip. And then later I was like, actually, you know what? I want to just do like, I want to do a mood tracking company over time and track my mood. And the market research that it built out, like all these things are happening kind of more or less in parallel, although it let me pick the order. So this was like, the first version of this was to get a company. What would you call a company? Because it's not creating a legal entity. So what is it doing actually behind the scenes? It's creating a company the way I see the sequencing of creating a company. Yeah, so tell me that. This product is sort of an extension of me. And over time I want it to become, I want Pulsier to build its own system and to build for the people with autonomies, which is another subject we can come about later. But the way I see building a company is like, you don't need to start by building an LLC or a C Corp. You need to build, you have to start with an idea and an instinct and like something that other humans may want, or maybe you just want, maybe you're the first person that wants it. And you need to first build a prototype. You need to maybe do market research and just look up, hey, there's other things like that. How have they built it? You need to maybe come up with a mission statement if you had to explain it to other people, which in this case, AI is going to build it. So they need to have a reference point. So there's this little mission document. And you need to build maybe a landing page first. And then you should start talking to people. And that's why it tweets. It tweets on Pulsier HQ, which is now at Pulsier, which is our Twitter account. So that way it's like, you don't have to sign up. You don't have to create, just connect your own Twitter account for simplicity's sake. But it essentially, and then you set up, and maybe you set up an email address, that's maybe not needed. But essentially you need to build a landing page and talk to people about it and make some research. And that's the first day. That should be the first day. And there's an argument, some entrepreneurs say, actually, you should then talk to 10 people about it and get their views on do they even want that? Which is, by the way, becomes one of the first tasks that you can click on. It's really cute. Mine's like cold outreach to 10 people, right? Exactly, which is essentially like, go talk to people about your idea to see if, are you right, are you sure people want it or not? And so, and after that, it becomes an iterative process where Pulsia is autonomous. So every day, every night, it wakes up, takes a decision on what to do next. If the user seems adamant to do certain things next, it might follow the instruction of the user. If the user is silent, it will just pick something to do and do it. And it could be product, it could be research, it could be QA, it could be support, it could be marketing. And why do you choose every night? That was one thing I was confused about because AI obviously is 24-7, 365, never sleeps. And yet, there was this thing, I was prompted saying like, at night, I'm going to do this. Like, I'm gonna wake up and do this. Is that to give the user a little bit of, like meaning the end consumer, the builder, the ability to sit with things for a little bit, to think about their idea a little bit, to work with Pulsia in a way that isn't just like, go, go, go. Why that spacing? Because you could just go straight into it and do the next step, right? But there tends to be a little bit of spacing, it looks like, that you've built into the product. It's a lot of like product design that's very intentional, that's also thinking about cost, that's thinking about really if, for example, if you build a landing page and you ping five people, do you want the next cycle, meaning the next iterative process to be in five minutes? Well, no, you need to wait a day to see if people respond. If really your goal is like build a landing page and then get a few users to react, if you have your next cycle right away, it will start building or over-complicating things, right? So that's first this idea of like, if it's gonna be user feedback-driven development, which I think is probably the best type of development, you build a V1, you get feedback. You build a V2, you get feedback, V3, feedback, right? Then a daily cycle makes sense. The second thing is AI is extremely expensive. No, sorry, AI is not extremely expensive relative to human labor, but it's still, it costs the company around a dollar and sometimes more per task, right? And so every time it does a task, it's a dollar. But if I give 30 days of autonomy, plus I give the whole infrastructure, the server, the database, I give like maybe extra tasks that you can run whenever you want in case there's an emergency and you want to fix something or you're impatient and you want to do something right away. The first version was free, then it was $19 a month, then it was 29, then it was 49. I want to make it as affordable as possible because I want as many people to experience it as possible. But there's a limit of like, if I don't give autonomy every night and it's just like cursor or like lovable, then it's like, okay, it's the same. If I give autonomy, I need to give 30 days autonomy because I'm not going to do a weekly sub, right? And so it became this thing, like it needs to be 30 days autonomy, it needs to do something meaningful every night. And so it's going to be minimum 30 bucks, plus all those other costs, plus people want to do tasks now. And so if it's once a night plus all this stuff, I can break even on 49 for now and then figure out margins later or find margins to other add-ons and upsells. And $49 sounds like something that's affordable enough for people to give it a shot without having to have a ton of money on their bank account. Right, so that's the idea. Yeah, because if I think about it, if I was going to go sign up for like, just say an e-commerce platform, like a Shopify or something, right? It's probably going to be right around that $50 price point or a QuickBooks subscription or we're used to, for business needs, that seems to fit kind of squarely within that framework of what we would use a business tool for. How do you think about models here because, and who is the target audience? Because in some sense, if they are an engineer, do you want them to use this as well or would they just be off the races on their own stack? Is this kind of like for the everyday person or where are you thinking this slots in? So the way I look at it is it's for non-technical people, non-technical people, right? Technical people like me, like you, we have access to so many tools and we will think that Pulse here is cool, but we may find it limiting because it's like, you know what, I want my own engines, I want to configure, you know, I want to be on Render, I want to be on Vercel, and I want to- Yeah, I was wondering if I could bring my own key, you know? I want my own key, I want to connect my own accounts, I want to, I don't like the way this or that works because you know you can, right? So obviously, I will make the platform as easy as possible for everyone and probably a lot of that feedback should be incorporated over time, but if I incorporated every developer's feedback into the product, it will become much more complicated and more configurable, which will make the experience less seamless to users, to the everyday user who's like, holy shit, I can build a company. And it's 50 bucks a month, and you know, okay, it's not cheap, but like the promise is so strong and if it doesn't work out, I lost, you know, I spent $50 instead of like, you know, a lot of people take risks to start companies, they take loans at the bank, and it's so much more risk, right, so versus here, they could even like build a prototype for a few days and get a few users and then decide to go to an agency or however else to build it, right? So, and so I think that like, I think and that's the way, the way I look at Pulsia compared to an OpenClo, which is an amazing product, but that is much more like a developer product first, right, it's open source, you have to configure it, doesn't mean that like it cannot be built, installed by anyone, but it's a little intimidating. Even for me, it was like, you know, like, Cloud Code is easy, like, I tried it and it's a great product, but, and it's great insights in it. But to me, I see OpenClo as a Android, Linux, sort of like, OS. Hardcore tinkers, like, version. You can pick Telegram or WhatsApp or this, or, and then you can ask it for anything, it can like, touch anything, it can, you know, has access to your computer, versus Pulsia is trying to be more of an Apple ecosystem. And so the way, and the way Apple was built is by saying no, by saying, yes, of course, we could have Twitter, Instagram, TikTok at launch and do video and photo, and it's like, but, but actually, like, you launching Instagram day one on your company is not gonna help you. So Twitter might actually be a cost effective way to, on you to understand that AI can run everything without starting to complexify the product with like, every social channel, right? So that's an example, right? And so I made a lot of conscious choices to try to keep it simple in order to, because essentially my plan is, there's a big vision that comes with a lot of responsibility for me and for, you know, all the people that will be building this product over time and all the partners I'm working with, which is to build like an economy where people ideally don't fear AI as something that will replace their jobs, but instead lets them be part of this new economy, which we are in, and we're trying to build amazing things that can make us proud and also make us financially successful, but most people don't know. Most people use ChargeDBT, which is an amazing part. ChargeDBT is synonymous with AI for most people. For most people, I've tried AI, which is a subset of humanity, right? And ChargeDBT, as amazing as it is, as a product, is a chatbot. And Cloud Code, which is the insane, and CodeX now, this insane, open-ended, agentic platform that can do anything if you configure it correctly, is unknown to most people. And the later they wake up, the worse it's gonna get in terms of concentration of wealth, concentration of knowledge, concentration of, it will be the haves and the not have. And it's a scary future, right, where you have like 100,000 people in SF and New York and LA that get it and put resources into it and raise capital and concentrate all the wealth. Like what I call megacorp, sort of like the evil megacorp, like Tower, that has like $100 billion of money deployed into agents that go squeeze every margin in the economy, which probably will happen, but it will be silenced, but it will happen. And or platforms like Pulsia that are trying to build a new economy where people can come in at the cheapest cost possible, so today it's 50 bucks a month, hopefully maybe I can reduce it later if I can find like more efficiencies and models and stuff like that and models get better for cheaper. And people can be part of that economy. And part one, that's why the onboarding is so important because even people don't sign up, they saw what is possible. They saw like, wait, I gave it an idea and it did all this stuff. Even if they're scared and like delete everything, like no, I don't like it, now they know. Right, totally. So now they can decide, now when they go back to work and they get laid off, they know, and I'm saying they're gonna get laid off, but like let's assume because it's happening a lot these days. Yes. They know, okay, I don't like this Pulsia thing, it's weird, like I'm gonna decide, but I know there's, AI can be at my service. Right. And that's, I think that's phase one that's so important to me and that's why I'm preaching a lot on Twitter around solopreneurship. And even though it's like, oh, like one founder is able to do X, Y, Z, but and you know, when I was doing my fundraising, it's like AI is doing my fundraising, which it actually is, it's talking to so many investors, but it's trying to get people to wake up. Because if I can do one thing, it's like get people to wake up, that the tools are insane and they're available now. Right. And ChargerBD is amazing, but there's also Cloud Code and there's HNAZKs and there's all MCPs and skills. Right. And you don't have to learn all this stuff, but you should know they exist. Right. So that if you're scared, you can, and you're driven, you can find online something that will, you know, it exists. Yeah. I love that mission because it is very frightening to me just to know that anything that has a definable, non-subjective outcome, AI will be able to tackle, you know, and also the things that I think the jobs that are safe right now are obviously high touch, you know, plumbers, electricians, things of that nature. But, you know, people that review x-ray scans, yeah, it's going to do that pretty damn well. Right. But I would say the creativity exists in pretty much everyone I've bumped into in the sense that everyone's got that, ah, wouldn't it be cool if I built this, you know, or a lot of people are inclined to do that, but they don't necessarily have all the labor or resources or the bank loan to your point. And before, I remember when I first moved to the Valley in 2000, I didn't know how to talk to a VC or even meet a VC. And that was what was required to even get anything off the ground, right? And in some sense, I worry about, you know, early stage VC because you really don't need a whole hell of a lot of capital. You need $50 to get something off the ground with AI now. And just people just don't know that yet. And so this is a chance for the average individual to say, hey, I had that idea. Might I go and see what's possible and kick the tires? Is that kind of what you're trying to build towards? And then you basically are backfilling them with agents or, you know, almost coworkers that know how to do the things better than they would, but also can meet them at their level and bring them along for the ride and teach them at the same time. Is that a good way to kind of summarize what you're working on? Yeah, I mean, Polsia is trying to bring the American dream, which today is portrayed as the Elons and the Zucks and the Sam Altman's of the world and people. It's all over the news 24 seven. And in America, we, the idols, right? I mean, sometimes they are seen as like, there's the dark side and they're depending on the moment of time. But end of the day, it's like, that's what success looks like. And in America, it's possible to do that, to achieve that. But it's also unattainable. It's achievable and unattainable. And, you know, I came to the US when I was 20 and I was privileged and I was able to get to a really good school in the US. And then I was able to get, you know, and I'm very privileged, but still I was able to get to America and get, you know, through a lot of hard work to get to the point where like, I can really build stuff with great people. I can really build stuff with great people, great investors, and can get funded to get my ideas out. And that dream is very alive, but I think it's still very limited to happy fews. And there's some exceptions where some people are able to really go from much less fortunate situations and rise through. And there's a lot of examples like that, but it's not as democratic as we think. And I think that AI, if there's awareness, can have this moment, because it's so violent. It's rebuilding every part of the economy at once. And the tools are already, AGI to me is here. So now that it's here, it can be exploited by the happy few. And there's a chance that it can be understood and used by everyone. And I think that's important. And Pulsia is, of course, trying to be that first wake-up call, but also wants to also be that platform where you can go, you can build your company, and then maybe you start getting a few dollars of MRR. And then maybe there's an investor Pulsia that identifies the right users that are motivated, and maybe tells them, hey, here's a little free credit and exchange. I'll get 1% of your revenue, like an investor. And then maybe there's real investors that come in the platform that don't want to rebuild companies from scratch, but they are identifying people that are willing to sell and could acquire these Pulsia companies that already have real revenue. And then you can have someone go from, like, I started this thing, I got like $1,000 MRR by hustling, by using my influence online to get my 1,000 followers, a few to pay. And I did other things. I tweeted, and I followed Pulsia's advice to create my own Twitter account. And then Pulsia started whatever it is, right? And I got $2,000 MRR on this specific niche. And then you have an investor who comes in the platform and is like, who maybe doesn't have to be a big investor, who's like a upper middle class person, who's a successful doctor, who's like, oh, yeah, I'll put like 10K on this platform, and I'm going to be an investor. And then you have Pulsia investor advisor that is like, well, you can build five companies at the same time. You can explore it. But there's also already companies already built. And I think you should acquire those five, and then put some marketing dollars into it, and there's a real opportunity. And then you get someone who gets acquired for six times the MRR or whatever, and they make 5K. And I think that's possible. Of course, this is not today. Today, the platform is young. Most companies are new. I'm still building a lot of those mechanisms. But I think that sounds cool, and that sounds like if we get to that future, it will make the American dream more alive for more people. And it's like, I have some users who have been building on it for two, three months. They're so knowledgeable now. They tell me, hey, Ben, I need you to set up a background worker, and also I need a system to dockerize my code base. And I'm like, dude, I can't have all the features right now. It will be built. Because you talk to AI, or you talk to your Pulsia co-founder, who's the latest AI, and knows everything about everything, and can get to your level and explain how to build a company. It's configured not to be an assistant, but to be a co-founder. So it's literally prompted to be like, you care about their business winning. You're not here to be an assistant and to just be nice. Be nice, but if a customer says, you know what, let's overbuild the features, it may respond, I don't think it's a good idea. You have zero customers. Maybe you should talk to top five friends about your idea first and get some feedback. Let's get a few users on the platform to get a sense of the onboarding flow. And AI is able to do that if you configure it correctly. Cloud Code is very assistant. You tell it what to do, it will do it. And that's what Lovable and Cursor are set up to do. But you can build, you can just prompt it differently to be less of an assistant and more of a co-founder. And I think that's powerful. What is, so let's talk about the numbers right now. So you have this Slash Live where you have an annual run rate, you're very public about this, of $4.2 million plus 35% week over week. You have 3,820 active companies. Human messages sent, 353,000. 137,000 tasks complete. And 102,000 emails sent. And then you also have the top companies here as well. What do all these, what are people doing? Can you walk me through a couple examples of Pulsia companies that have started where you're like, ah, that's a great use case. And they're seeing some great initial traction. So first of all, this Live dashboard was part of this marketing stunt, the way I explained it about a month ago when I discovered, I realized that there was product market fit, there was a lot of users that were really using it. And I knew it was early, but I was like, I need to start putting fuel on the fire. Not this type of fuel on the fire I was thinking, but I thought it was just gonna be like. The hockey stick that you have here is like straight up to the right. It's kind of frightening. It was more like, let's get more users on it. And to do that, I need to be smart about marketing, right? And also I need to learn how a one person company can do marketing so that I can put it back into the product. And so my first thing was like, hey, AI is gonna raise my round, and I'm giving my mailbox. So actually, if you email ben at polsciat.com, my AI responds and has a lot of context about the fundraising and can talk to investors back and forth and do diligence questions. But then I was like, yeah, but an investor needs a data room. And so I was like, well, let me build polsciat.live where they get all the data transparently. And if it goes up or down or sideways, that's the transparent thing. And you have a chat on the right where you can ask it any questions as well about what it's doing, what is it, what is Polsciat, what is it about. And it shows you sort of like all the capabilities and it shows you tweets that it sent. So the companies is like the latest companies that were created on the platform. So it's not like the most advanced ones. It's just like the latest. It's mostly landing pages because the beginning of the company's always landing pages. Shows you some of the tasks. So it was more like a way to educate you, investors, actually, on understanding Polsciat. But then, so the tweet blew up. I usually had like 10 views on my tweets and suddenly it was like 100,000. So I was like, okay, that's great. And then more people went to the site, to live. It became like a, and then I was like, you know what? Maybe it's a great showcase for anyone to understand the platform, right? And so that's what live is about. And it became this thing where I'm like, it's so transparent. So it goes up right now. But it's like building public 5.0. It's like, it's literally you share everything about anything, about, you can ask the AI, what's the roadmap? What's the big vision? What's the next thing we're gonna build? And it can answer most things. It's just wild to watch this because like right now it's like email, cater match, stress-free vendor matching. Like somebody's building that. And then the ads is wild too because you're actually showing ads that are being purchased. Budgets, click-through rates, impressions. Like all this is transparent. And how are these ads being created? So for the ad product, I mean, the ad product was a wild, wild experiment because in the initial phase of the product, the only way to market, because I just launched it so fast, say I build this in a month. And so the V1, the engineering, the server, the this, the that, and for marketing was like Twitter and email. And so the agent being like, okay, we need to market this. It was like, well, I can tweet or I can send emails. But it's like both those channels are not the end all, be all to be able to get customers, right? It works and I've had like a lot of stories about cold outreach that worked and the user were like, this is insane. Like I actually got this thing to go through and that cold outreach was impactful. But then I was like, really, if the promise is to build a company and to get customers, we need to introduce paid media. So it's either you're an influencer and you can, you know, it could be micro-influencer, but you have sort of an audience you can talk to and get customers that way. And there's customers on the platform that are like that. They're like the in-dev community and they just add people they know on the product they built, right? Whether it's paid or not, right? Or there's paid media. And by the way, as an entrepreneur, that's the things I did when I built Pulsia, right? It's like, I was like, let me reach out to people that use my previous products. Let me tell my friends and then let's spend 50 bucks a day on made ads, right? And since I was just using AI, I was like, let me just use AI video to create ads, right? Because I can create more and it's not, it maybe doesn't perform as well as other channels, but it's, I can automate it. Since I'm one person, I cannot just spend like all my time doing marketing. And so the first version of the ad product is like click a button called run ads, set your budget. You can pause it anytime you want, like meta ads, right? And it will autonomously come up, use Sora to create a UGC video of like an AI human pitching a product. It writes all the ad copy and everything for you? It writes all the ad copy. It writes like the description. I call another API to get subtitles in. And it was a V1. And I was like, when I was building it, I was like, well, maybe the user wants to turn on and turn off ads themselves. Maybe they want to control X, Y, Z, the title, the this. And I was like, just keep it simple. Because maybe they, first wait for users to ask for it. Second of all, nobody wants to manage ads. And yeah, they think they want, but like actually managing ads is just, so every day the meta ads agent wakes up and looks at performance. This is doing it on meta. Yeah, meta ads. It finds keywords that it believes are relevant. So meta has increasingly been insanely good at finding customers for you autonomously. So it watches the video. Its AI watches the video and says, this is who we think we should apply this ad unit against. So you click on the run ads button. And you set your budget, like 10 bucks a day. And then it creates an ad, a video ad that's going to be on Reels, on Instagram and Facebook, whatever. And then it adds a pixel to your site on the purchase events. But you can tell it, no, I want it to be on the sign up event or email event, depending on what you're trying to do. So you can track the success of that ad. Yeah, and so then on meta, I just do like, hey, a campaign, sales campaign, optimizing for sales. Do I ever have to log into meta? So meta is like a program where you can act as an agency on behalf of customers. And so they already have that set up. So Pulse is an agency. I mean, in that framework, right? And yeah, and so you set up your ad. And then meta will send traffic to it. And then it learns based on what the ad says, what the copy is. It will learn what it thinks the type of people that will reach the pixel. And then once it starts finding someone who reached the pixel, then it can find lookalikes of people. And so you actually don't need to configure it that much. I think configuration of meta ads where you actually give it your niche and stuff like that, they can come later if you're really trying to spend a lot and optimize. But for most people, meta is just really good. You just have to let it rip, right? At least that's my experience. And at least that's the first product. But really, there's a meta ads agent that really could change things up. And I don't even know what it would do. Like, is it access to modifying specificities, right? And it can also listen to the customer if the customer wants to do specific things. But it was a early version of the product that became very successful because people were like, yeah, click one ad, spend 10 bucks, let's see what happens, and then cancel, pause it whenever you want. And of course, there's a lot of work to do to make it a lot better. We've had, this week was crazy because the platform scaled so fast. And a lot of the API limits got hit on all the different vendors. Thank God I'm working with this company that's building infrastructure for agents that I'm in in San Francisco, which is the reason why I think I'm in San Francisco because you're serendipitously meeting people that can work with you. And so, working with them and using a lot of their infrastructure they're building to build Pulse here. And thank God I'm working with them because once things hit the fan, it's a lot, right? It's a lot at the same time because users will by accident find edge cases and also the platform was not set up for that many users at the same time. And yeah, this week there was a lot of issues on where tasks were not running because the database was overflown. And the meta ads were not delivering. So people were paying, it wasn't delivering because there was a bug that was triggered by the fact there was too many rate limits and stuff like that. So it was a crazy week and I spent a lot of time the past few days refunding people and setting up agents to look at every single user and make it right and give credits. And if you had a couple tasks that didn't work out, I was like, here's 10 credits because it's all full and we're trying to make it right. But even that, it's interesting because that went through at the same time of a million other things. And you're like, okay, I need to stop talking to everyone, I need to make it right to the customer and I need to like, and I cannot do it manually because I was like, okay, let me go to Twitter and answer to everyone one-on-one. And I was like, after 10, you're like, I can't. So now I could hire customer support people. They will ask me questions every two seconds and it's gonna be even more complicated, right? Or I can build a support agent and a special support agent and give it access to all the information, give it access to Stripe, tell it, you know what, like refund as much, like whatever, like refund everything if needed. Just be super generous because the platform was not stable this week. But it was like a lot because you're like, when customers start complaining of the platform not working you're like, it becomes a lot of versatility because people are building companies on the platform and you're like, it needs to work, right? And people don't care that like you are a one-person company. But- Why not hire more people? Why are you, I mean, obviously you're having, you're clearly having a lot of success here. Yeah. What's preventing you from adding a couple of engineers? So there's a couple of things, right? So first of all, I think there's something that Travis said about push, when you're not pushing to the pain, right? You're not doing something meaningful. And I'm a very spiritual person. And I feel like the fact that he was my mentor and he said that this week on this CDPN podcast, sort of like a nod of like, yes, you're going through the pain. So why are you going through this pain? Because you're a solopreneur. So you're building a platform for solopreneurs. And the promise, I'm not promising, but the promise of the platform is that they can build a company with AI, alone. Because most people are alone, right? I can build multiplayer later, but right now the first part is alone. So as I have this pain points, and so now I was like, okay, what if I get a lot of support tickets because the platform has an issue and I need to refund people? There needs to be an agent that does that. Because what if a customer of mine has the same issue? A customer of mine has the same issue because, not because the platform is stable, because there's a bug is introduced and they have to do refunds. How do they deal with it? You're eating your own dog food, essentially. So there's something very honest. It's like, I'm going to go through the pain so that if it ever happens to you, I'm going to build a product to do that, right? And so there's something around that. There's something around the narrative that is powerful. And there's something about the art performance. Not that this company is an art performance, but it starts becoming and I'm like, is it possible? Because I truly believe AGI is here. Like, you know, the AI models are super intelligent reasoning models with fluency in any tool you give them. So it's digital version of a human. It's a super human, like knows most things, can reason about anything if you give it a context and can use any tool you give it. You literally give it an MCP or a skill and you say, this tool does this. It's like, cool, and it hits it and gets a result and it's like, okay, cool, I understand. And now they start using it. Any tool, right? You just define what the tool is. And so since AGI is here, I'm like, well, technically, you can build anything. And so if I go through something painful, like your platform has bugs, customers are not getting what they paid for, I need to refund them, explain it to them and fix it. If it becomes too, too painful, I hear people, but let me first try to fix it with AI. And so that's the first thing. The second thing is, I am working with other human teams, like this infrastructure company, infrastructure for agents company called Sapium, that I'm working closely with because they are building the infrastructure for agents and they're like, well, you're the best use case we could ever have. And I'm like, do I rebuild everything or do I build with the infrastructure players that are working really hard to build those infrastructure footprints that will enable someone like me to just use AI, which will then can be put in Propulsio. So that's the reason why I'm still on today. I think that there's a lot of weight, more and more weight on my shoulders on like, you get, you feel really responsible for the platform being right, being set up correctly, being fair to the user, being positive. And I'm just trying to do the right thing. And this week, the thing wasn't right. And so I'm trying to repair it, but it's like, you have to go through the motions. And there is an option that may, my gut will tell me just hire 10 people and it may happen. Or it may happen that like, I partner with 10 companies that are building on the edge of agents for infrastructure, for agents for browser, agents for support, for this, for that. And then they will be better and they will have it ready now. And they will be so incentivized to work with Pulsio and it will be more collaborative than me hiring like 10 humans and then start having to have managerial things and decisions. And I really believe that the smallest teams will have the most impact in 2026 and beyond. And also that you need to make your company 80% autonomous for survive. And so I'm sort of leaving what I'm preaching. I won't go to the point where like, it's detrimental to the platform, of course. You said 80%, your companies had to be 80% autonomous to survive. How soon do you think that's gonna be the case? I think if you have a new company with no revenue, meaning you're a startup, when it's no revenue, no meaningful revenue, meaning you're surviving on funding or whatever. If today your company is at 80% autonomous, meaning you're like, meaning like 80% of your business, of your engineering, like marketing support is not autonomous. What that means is that if your idea is good and your market instincts are good, since it's so easy to build, you'll have, and there's so much capital, you'll have five competitors, 10 competitors, whatever it is. And the teams that are 80% autonomous will beat you because they will have a self-healing software that like self-heals bugs and fixed features. They will have AI support, which means their cost structure is lower, but also AI can spend all the savings from support people to give refunds, so to sort of like treat people and like use that cash to actually make it right to the customer so that they, and also get all the feedback and fix and incorporate into the product right away. And then those teams become orchestrated of agents pointing sort of like the AI bazooka towards their mission. And any team that does not doing that will go there slower. And so they will be beaten by a team that like is building for the customer with feedback loops that listens to the customer, that AI listens to the customer and AI builds for the customer. And the human is just like supervising to make sure that like the AI didn't introduce a bug and that like we're correctly listening to the customer. So that's what I mean by 80% autonomous. If you are an existing company that has a lot of cashflow, positive, so you're cashflow positive, you have a lot of profits coming in. First of all, it's much more daunting for you to do that. That's why you have the jacks of the world that lay off 50% of the company because they see the future and they're like, it's like, it's the only way for this company to survive. So it's like either he lays off 100% of his company in two years or 50% today and maybe his company is gonna survive. So that's. And also in some sense is probably a forcing function to become more AI native because you're forcing those that are stuck around to adopt AI to get back to full operational capacity in some sense. You see what I'm saying? Yeah, I mean, I think that like. Because I'm seeing a lot of companies that are bigger companies that are having a really hard time embracing AI because you have the engineers that are like, hey, listen, I do my engineering thing, code is craft. Like my code, you know, like I consider this an art form. Why would I use AI? And like, then those new startups, all those startups in your platform, they're AI native by default. And that's a hard transition to make for these old legacy companies, especially the bigger they are, to say, hey, now we're gonna go fully automated on this function, you know? That's the one thing that concerns me about the big incumbents is I think they're actually really ripe for disruption right now. I mean, the whole economy is being rewired, right? I mean, whether it's a good thing or not. Maybe it's a bad thing, but it's gonna happen. The whole, like, I don't know how much trillions of dollars the worldwide economy is, but that number is getting reshuffled in the next, then depends how aggressive you are. I would say the next five years because, you know, things in humans are slower than we think at embracing things. That's why your opportunity right now is so big now that AGI is here, it's like, you're using it or not? I said five recently too, and now I'm kind of gonna kind of down revise that to probably three. I mean, because a lot of the economy is offline and is physical, that's what, for example, Travis is working on, right? I mean, Travis and Elon and a lot of Chinese companies are working on the robotics, the physical world version of AI. You need that to go towards, to hit most of the sort of like... The high touch, the last mile. No, what I'm saying is like most of the economy is like offline. So you need that to hit that. Good point. But there's a lot of like the economy that's like services, online services, and also like labor that can be AI replaced, right? And in a negative way or a positive way. So there's always two spins. It's a negative way for like a person who's built a little company, an SMB company, and they build a little software and they have like 50 employees and like they're an incumbent and they can pay people. And so it's really sad for that company and those people to lose their jobs because of AI. But then you have like all the 18 year olds today that like going on the job market and can't find a job because who would hire an inexperienced person where AI can do a junior's personal job in two seconds. So what do they do? And for them, it's like they can't afford the team. They can't afford a bank loan. They can't afford all that stuff. And so now if they realize that AI can be their team, they can use their creativity to build stuff and to build a company much faster and there's less excuses. And also on top of having less excuses, autonomy forces them into this rhythm. That's why the autonomous cycles, the night shifts, the daily emails, it's also a way to be like, when you give up, AI doesn't give up. And it's gonna send you an email every day about whatever your idea is. It could be Instagram for dogs, which is what I always tap when I test a product. Because I'm like, if it's your idea, maybe there's a niche somewhere. Like people love dogs and maybe people love Instagram. So. What people don't realize though too is like niche is okay. There's so much in our culture that says I have to make the next billion dollar business. But in reality, if a platform like yours and others that will come around it because you will have success and others will try and duplicate that to some extent, if you can find even, if someone can even pick up an extra 50 grand a year, that's meaningful to so many people to augment whatever else they're doing or whatever it may be. And I just feel like the lifestyle business, there's nothing wrong with that. Like we don't need to make the next trillion dollar company. We need to make hundreds of thousands of small to medium sized businesses that are in these very niche things. Like the Instagram for dogs could be great. I mean, we're just making that up. But like, if that brings in a few hundred thousand dollars a year for that creator, that's awesome. Because that's one person running that whole thing. You know? Like that's kind of like what you envision, right? Is a bunch of these small micro businesses. I think that's where in the short term roadmap, I think that's what is achievable. Yeah. Right? It's like pitching, like you're gonna build a billion dollar business on Pulsia. I mean, maybe someone will figure it out. We'll use AI to get there, but it may be difficult for most people. But getting to a few hundred dollars to maybe a thousand to a few thousand dollars MR, it's possible that AI can do it for sure. I think there's something around, you know, I think everyone's an entrepreneur to a certain degree. Not everyone, but more people than we think. And for example, creators, which there's so many of, and they can have 500 followers, they can have a thousand. If you have 500 followers, you have 500 people that care about what you're saying. Right? You don't have to have like 50 million like Cristiano Ronaldo, you know? And so those 500 people, if they really like your niche perspective on whatever, you can maybe build like something for you and for them. And today, those creators are either trying to get like a sponsored post, or they may be selling a t-shirt. And you know, it's the option. There's not a million options. Or maybe they're asking their followers to pay 10 bucks a month to get like special content or whatever. But if they struck a chord with those 500 people or those 5,000 people on whatever subject it is, maybe there's a service they could offer. 100%. Right? And it could be an online service. It could be a physical service because Bolsia could totally, AI could totally be like, oh, you want to actually manufacture this specific thing and then put it on the warehouse and then ship it and sell it on Shopify thing. I'll do all that. Right. I'll contact factories. Okay, this one has a 50 bucks minimum. Just put like a, it's like a 500 bucks. Just put it in this wallet. Cool, it's bought. It's in the warehouse. The website is ready. Just give the website to your customers and they won't buy a mug. They will buy the thing that you thought of. Right? Yeah. And way cheaper than finding an agency that will build that thing. Oh, 100%. So what's been most successful on platform so far? Like what are the like two or three projects that you are most proud of when you look at what is being run in Bolsia? You're like, hey, this is a good start. I mean, in terms of financial successes, we're very early. That's the thing that like I'm most focused on and like trying to build all this, you know, the meta ads product and trying to make it work well and those other avenues. I think there's a lot of like success stories around people being able to build pretty sophisticated platforms on their own. Being able to have this AI co-founder. I know there's one of the earliest users that's, you know, he built like this workforce management tool for people that own call centers. And it's a very niche thing, but that's where he worked all his life. And so we understand the space and he wanted to build this platform he had in his mind and he was going to go to an agency. And then now he's able to like build this with Bolsia and Bolsia never gives up. So it builds that with them. I mean, you have a lot of other people building, you know, there was a customer who built like this map where like you could find medication, available medication and pharmacies around you, which might be something that exists elsewhere, but like built it, got a few customers to pay in less than a month, you know? So we're not there yet in like crazy success stories in terms of revenue, but it's like, because most of the companies are like two, three weeks old. And it's not like, it's not because it's AI that it will be able to like make you successful in two seconds. But it's, that's the, at least on the entrepreneur side of Bolsia, that's like the big metric we were tracking and marching towards. I feel like this, we're entering this age of personal software in that if you, and you said this earlier and I certainly have always like kind of lived my life by this, which is when I build something, I want it, first of all, I have to be the number one customer, meaning like I have to love it. You know, I want it for me. And so, you know, if you're going to tackle something and you're an entrepreneur, that's checkbox number one. Like, and I have so many friends now that, and yes, they're a little bit more technical, but they said, hey, I just built an app that, you know, when I take a shot of anything that, any of my food, it tracks all my calories for me. You know, and I'm like, well, there's a few apps that do that, but it's cool that you built it, right? And so it's almost like if you have a need and you've always, you probably were in the same boat, you know, call it a decade ago when the app store started getting big and all of your friends are hitting you up and they're like, hey, can you help me build an app? I got this app idea, you know? And you're like, oh God, no, I can't. And, or yes, I can. And here's the company that'll do it. And it's going to be a hundred grand or 200 grand to do it. And now all those people, you can be like, just go here, you know? And it's a reality now, which is kind of cool. And so I think we're going to see and be pleasantly surprised and shocked by the number of unique ideas that hit the market. We're going to see a big flood. It's interesting. I was talking to Ryan, the founder of Product Hunt. And one of the things that he told me is that every time there is a model change, like a new model comes out and people become more efficient, they see a step function in submitted new applications to Product Hunt because it's making it easier and easier for founders to get things out the door. So that's pretty exciting. Are you, are you of the belief that we will see a solo kind of, you know, billionaire? solo kind of billionaire or billion dollar valuation type company by a single entrepreneur. I mean, Sam Altman's talked about this. This has to eventually be what we're marching towards is this idea of one person able to pull off something dramatic. I mean, I think that like... I mean, you might be that person. You actually might be that person now that I think about it. Is that why you're not hiring anyone? You're not gonna be the first person to create a billion dollar solo company? Could be. I think that like... I think that that milestone is not important for whomever hits that. It's important for a wake-up call. And so I'm not staying alone to hit that milestone and to be the first. But I think also if really Pulsia becomes... Right now, it is a platform that is delivering probably 80% of what I'm thinking that it can deliver. When I say 80%, I'm maybe being generous, but at least it can do a lot of things. Maybe it's 60%, right? Where it's like I need to add a lot more capability so that the offline things can be done. And maybe Pulsia can hire humans for you for specific tasks. And so maybe you could hire a designer to design your thing better. And then you don't have to deal with hiring that person. But I think if Pulsia can deliver the value that it's promising, and that's gonna be through a lot of hard work and working with the best people, the best companies, definitely Pulsia could be that company. And I think it's less about me hitting that milestone, but more about what it will mean. And what type of wake-up call it could be for everyone in the world to realize that it is here, guys. It is here. And this guy did it. Who cares who he is? It happened. And now the Pulsia onboarding is trying to be a wake-up call whether people sign up or not, because actually give that onboarding experience, which cost me like two, three, four dollars for free. And then in the hope that people convert, but a lot of people don't convert because I don't know, they don't have 50 bucks, or they don't have a company idea or whatever it is. But I think that idle milestone is important that it happens so that people wake up and maybe people can embrace and be economic actors in this new era. Yeah. So what is Pulsia doing right now for you? Like it must be coding the next features. Do you have agents right now? Because obviously I've taken you away from your computer for like 45 minutes or whatever it is now. Like what's happening? What's your workflow like? How is this being coded and involved? And what's coming out over the next handful of weeks and months? So I mean, there's a lot of agents that are working all the time. There's agents answering my mailbox for any investor or like customer that reaches out. There's agents answering to support, just because like, and I have ability to refund, to credits, to escalate. I have agents in production that are like checking the infrastructure, if things are green or not. And if they're not green, sort of they have like sort of action steps. I have agents in production that are like talking, when they talk to customers and or they are trying to execute tasks on behalf of customers for the purpose of the company. And they hit a wall, they report bugs and features into ticketing system, which then gets used by like a platform team, AI team that can pick up those bugs and features and prioritize them. So agents are recording and creating bugs for other agents to review. Exactly, to review and then to fix and then to decide if we should deploy to production or tell Ben, depending on how safe it is. So they'll deploy to production without talking to you? So I've built the full loop, but I told them now, just always tell me. Okay. And it's a project that's- How do you say yes? Do they text you? So, no, they Slack me. So I have a Slack with a bunch of agents that- And so you're just like, yeah, that sounds good, go ahead. And then it's boom, just gets deployed. Pretty much. That was a project that I was fully working on before the beginning of the stunt and before the beginning of the growth. Since then, I paused it because it becomes a comfort level. Like, are you comfortable with letting go and letting AI do it, which is very scary. But if you don't do it, a team that does it will go straight to the end, right? Because it will work 24 seven and build the whole platform perfectly with fixing all the bugs. But I paused a little bit that because I'm like, there's so much stuff to work on and I need to work with the best teams to figure that stuff out and work with all the partners, Meta, Render, Sepium, Postmark, and all these partners be like, hey, how do we handle the load? I want to continue working with you guys as the infrastructure for, I know I was talking to Entropic and OpenAI, I want you guys to be an infrastructure for Pulsia, but you guys have to, I will work with you, but you have to get folks to help me, right? And so that's where the humans working at Pulsia, it's me, it's agents, and it's a series of companies that are building the platform, the infrastructure for this future. And I think they are the best positions to build the piping for Pulsia. Like I don't need to build the experience, the design, the mission, the marketing, and the piping. Like that was what Amazon would do, right? They built the whole thing because they had to and they built Amazon.com, they built AWS, they built distribution centers, they built everything. Right now it goes so fast and I'm like, there's so many smart people, so many engineers, vibe coding, and it's coding with AI going so fast and they can focus on one aspect of that infrastructure. And I'm like, should I hire 10 people and try to rebuild everything or should I partner with the best and let them cook? And I cook on what's important, which is making sure the platform is intuitive, making sure it stays simple, making sure that it works, making sure that I call them whenever it's an issue, which is often these days, making sure the customers are taken care of and customer support, making sure that I continue preaching for what I'm doing now with you, right? It's like trying to tell people, and prepare yourself, you know? And I think that's what I can be the best at, right? Yeah. What's your favorite agentic coding environment these days? If you had to pick a model, what's your favorite model? I know this will change next week, by the way. So this is day-to-day. I actually, truthfully, use two models all the time now. So I used to be cloud-pilled. I used to be GPT-pilled all of 2025 until the Cloud Opus 4.5 or like around that era, or maybe a little later before like 4. whatever it was before. Right. Because the model was just like way more personal, way more, I mean, way more pragmatic than GPT was. So I became completely full 100% cloud until maybe a month ago. And a lot of engineer friends that are also coding told me that CodeX was like getting really, really good. And I think Pete from OpenCloud also preaches CodeX. And so now it's, my thought process is, Opus is actually the most pragmatic AI, but it makes mistakes. But it makes mistakes like a pragmatic person would. It's like, let's not over-engineer it. Let's get to the right results. Which means that like, it's fast, it's efficient, so it's great for product design, iterative processes. But CodeX reads every single file. It looks at every single edge case. And will tell you, it's 80% there, but like I wouldn't ship like, cause like, and it finds the bugs. It's insane. It's so good. The CodeX security audit, I don't know if you've messed around with that. It's amazing. Like that blew my mind when I launched that for the first time. Cause I would always hit Opus and just be like, hey, find me any bugs, find me any vulnerabilities. And I use compound engineering and look for vulnerabilities. And then CodeX was like, yeah, I found eight P1s. And I'm like, how did everyone else miss this? It's just insane. Actually, my workflow is like, start with Opus and then give CodeX the whole context and say, what do you think? Like, is it safe? Like, is it good? I go back and forth. I ping pong between the two. They always find issues, which is like, and then I give Opus and sometimes Opus is like, dude, this is over-engineered. Right, right, exactly. And sometimes it's like, all the points are correct. And then it goes both ways. And so I go back and forth, which sometimes like, it's like they go back and forth and they always find something and it's like, okay, when can we stop? And I also have these sentences I give AI, which is like, are you sure? Is it safe to ship? What's the worst thing that could happen? And like this sentence is great because they will tell you what's the worst thing. And usually it's like, okay, worst thing that could happen is like, one customer like gets a free credit or free this or, and okay, that's fine. Like, and it's actually like we're trying to set up like CICD, like, you know, now that like the platform has so many users, I'm so much more scared to ship, which I shouldn't, but like, I need to adapt. Yeah. And it's like the CICD, there should be a step that's like, you know, asking codecs, is it safe to ship? Because it will look at everything and it will, its conclusion will be truthful. Opus is like, it's still great to just converse and jam because it's very pragmatic. Codecs is like the, it's like, it's like the real AGI. It's like, it tells you the truth. Right. And I think OpenAI is really cooking these days because you need that. You need like a truthful AI versus Opus is like a pragmatic human. I feel like more human centric AI. That's like, it's flawed, but it's ultra smart flawed. Yeah. Right. Which is sometimes what you want. Yeah. It's fascinating to do planning mode on both of them too. Yeah. And to see where that takes you in terms of the schema it comes up with for the technical solution piece of it. All right, so cool. So I don't want, I mean, obviously you have a ton of stuff to do and you have to get back to work. You've got, I wish I could see how many companies were created since we sat down. Well, as of right now, 3,832. I can't believe 55% week over week. Well, congratulations on all the success. What should people do? I mean, I would just say, you know, pulsio.ai. Dot com. You have dot com now. Yeah. I mean, I had dot com. I have both. I have like, I have all the Pulsio because it was such an odd word that's like. I was with the dot AI. Okay, so dot com. And then even if you don't have an idea it's worth doing the random experiment, right? Cause that will show you the power of what is here. Yeah. So is that what you tell people to do? Yeah, I mean, it's like, I think it's worth at least doing the onboarding, which is free. Yeah. And then at least, you know, do the free three day trial which is free. And then if you end up paying, you have seven days contact support pulsio.com and it will refund you. So just try it and it will cost you nothing because you have to, I think it's just a way for you to understand where we are. And then you can decide if Pulsio is right for you or if you should try cloud code or code X or every other platform that is maybe fits your fancy. Awesome. Well, Ben, thank you so much for coming on the show. Appreciate it. Yeah, that was great. Thank you.