It's gonna work every night on your behalf and then send you a recap on the morning at 8 a.m. Where did this come from? Where did the inspiration come from? And to this point now we're shitting out a routine new ARR. I mean inspiration came from like a decade of building to be honest. Like it's not, it's I think it's like the collective sum of all my experiences. Ben Zera, the founder of Polsia. Polsia is an AI that builds and runs companies autonomously. You give it an idea and it will go about building the product, fixing the bugs, snaking like a research, you can tweet, you can set up a company email and hit many more things you know. Hit 10 million in run rate with one employee yourself. Yeah. Ben, welcome to GTM now. Hello, how are you? I'm well thanks. How are you doing? Pretty good. They'll tired but are good. You've had a busy stretch. You just raised 30 million dollars that a 250 million dollar valuation. Big congratulations. The launch was pretty viral. I'm both ex-enlinked in. Many different factors to that. One is the name and where we can start. Polsia is AI Slop, spelled backwards. Tell us a bit about the naming decision. I mean, so the name came about when I decided to start a new company in April 2025. I got most of the year ago and I wanted to go all in on AI and build with AI and build. At the time, I had an idea for the first product was a NAB that would build apps. Since then, there's a bunch of other apps like that on the App Store but that was the initial concept. But I knew that product wouldn't be that or be all. And when my lawyer asked me, "Hey, you need to come up with the name for the C-Corp." Right? For the Delaware C-Corp. I was on my couch at 10 p.m. at Paris and I was like, "What about Polsia?" Like, "Yes, that sounds bad." It doesn't really matter the name of the C-Corp or whatever. That's the name. But then my investors would keep on referencing the company as Polsia. I was like, "No, Polsia is just a C-Corp name." Sometimes it doesn't really matter that name. My product is blanks. That's the in-and-product I sell. Fast forward, after building five different SaaS during 2025, that we're not really working. That product blanks didn't really take off. I was like, "You know what? I'm going to build this company OS." This OS that builds companies for me so that I can build many. Actually, it's a really good idea. And I think users, the consumers would love it. And so I think I should build this as a consumer product. I was like, "How should I name it?" And I was like, "You know what? That C-Corp is called Polsia. It's actually name I could get a.com for." It's a really good name, actually. And so I looked for the.com and I got it for a thousand bucks. It's a sign." Because a thousand bucks for a six-dollar.com is actually pretty cheap. So that's how it came about. That was no like a grandiose plan and anything. I think that it just felt right. And I think all of Polsia, the way I built Polsia, the design, the UI, the gamified UI, the vibes, the Polsia's personality. Everything came to me naturally and it just felt right. And I stopped caring about how people would think about how it looked or what the name was or what it did. I just thought that it's something I wanted to use myself. And I thought that people would want to use. But fast forward now when you have all these people freaking out about the name, I'm like, "This is the best marketing I could ever hope." Because I mean, when people complain about specific aspect of the product or like that if I could get it sloped, that's something else. But people just like repeatedly think like, "There's no way this company's called AESLA backwards." I'm like, "Okay, amazing." I'm getting a lot of views for free. Just the people having fun with it. So, and when I talk to my customers, which I actually just came out of like an interview with one of my customers, they never mentioned the name. That's not a never or something that they bring up, right? Like, "Oh, what's the cool boss here? I don't like the name." Or, "Oh, the design. Why don't you make it more. " So, like, there's never what to talk about. What do you talk about? It's like, "Polsia, did this for me? Like, Polsia is maybe an entrepreneur. I'm building this one thing and I really love. Is that a thing?" And that's what I care about. It's like, you know, what, if all customers were like, "Change a name, maybe I changed a name." But like for Twitter to be like, "Why did you call it like that?" I'm like, "I don't know. Why did Amazon call itself Amazon? Like, is it because it was in a forest or something? Why did Steve Jobs call it Apple? Because like, he likes fruit? Like, I don't know." Well, I think it was a marketing genius play because so many people are now referencing "Polsia by name." But less like you said from, "Does it matter or not?" It's just been a memorable marketing moment. Is what it's kind of created or seems like it's created now? No, and I think what matters is like in marketing is like, end of the day, what, you know, what matters is like, are people. Is it pushing people to try the product? And then, and then that's where the real game begins. It's like, do people like it? Do people stay? Do people retain? Does it add value to their lives? And that's what really, really matters. And that's like, what most of my time is focused on? 100%. As it always should be, I say, brand is so much more than a name, obviously. It's really how you make people feel, the value you add. And, you know, we talked about the name itself, but for anyone unfamiliar with policy, I'm sure many people saw the launch. But would you mind just explaining what policy is? Of course. Postia is an AI that builds and runs companies autonomously. You give it an idea, and it will go about building the product, fixing the bugs, making market research, in tweets, it can set up a company email and like, no fine leads, respond to support, run ads, and many more things, you know, just introduce the, you can buy custom domain for you and find a name and register it and like set up the DNS and everything. And so the idea is like, is that, you know, it works autonomously for you. So it will take decisions on your behalf, but you can also guide it. Remember what you say and then act on your behalf towards building the company that you idea. Strap is already set up for you. All the APIs are already set up for you. And so for you, it's really the idea is like, you just talk to it and you just give it ideas. It's going to work every night on your behalf and then send you a recap on the morning at 8 a.m. what it's going to summarize what it did, what worked, what didn't, what's the next step. And then you can reply that email onto the dashboard. So it's a very simple loop where you get an AI employee and AI team, an AI co-founder, however you want to call it. And it's going to build for you as a human would like if you hire an employee that gets going to work when you don't tell anything it might continue working, right? But if you go to it and micromanage it, then it will listen to you. And if you say, I trust you work on this mission, the employee will try their best. And so that's sort of like the way POSIA behaves. It's not perfect yet. It's like it has limitations. It doesn't have all the tools yet. But the idea is over time to give POSIA more skills, more tools, more more more ways to act on the behalf of the user so that any idea you have can be created. Whether it's like creating a new business, whether it's just trying out ideas, whether it's like a personal project, whether it's like having it help you on an existing business, which a lot of users are doing, it can do a lot of it. I love it. One of my favorite or I guess favorite quotes in that we want to disprove it is the best ideas and most ideas love in the graveyard and feels like POSIA is really positioned to actually stop that happening and just democratizing access to people building. Now for anyone wondering, there's so much noise around AI right now and so much opportunity for people to build why POSIA versus any other area that they can be building on for someone sitting here with an idea. Yeah, I mean, it's just interesting. I think POSIA at its very core is no difference than cloud, GPT, cloud code, codex and every other tool that exists in the sense that it's spirited by an AI that I can act and can do things. Why are it really the first? It's two main things. Number one, it's already setting up all the infrastructure and environment for you. So it's already picking like a hosting solution, picking where the code is going to be hosted. It picks the database provider. It picks like it sets up an email provider. It sets up a stripe account. So all this stuff that you would have to manually set up if you would start a new business, POSIA does it for you automatically and you doesn't ask you, if you ask cloud, okay, I want to host my website. It's going to say, well, you can use first cell, you can use HeyRocco, you can use AWS. For most people, I don't care. I just want it to be online, right? And so POSIA just puts it online and it doesn't really ask you where. It doesn't really matter. It's included in the subscription, right? Number two, the big difference is that it's autonomous, meaning it will work every day as long as you pay it. It will work every day on your idea and never gives up. And so that's a big difference because if you go to like, let's say a cloud, it only.
works if you prompt it. So that you add this whole prompt engineering thing. And so what does that mean is that I think a lot of people that I talk to that use policy are they've tried the AI before, they've tried other products, but then they tried it and then they kind of gave up because you know you get it really excited for a few days and then you're like and then you hit a bug or like you get a frustration and then you give up. And I think that policy is not going to give up on you. Like it's actually given an idea, it will continue on your idea as long as you decide to keep it running. And I think that's a big difference because then policy are prompts itself, right? So if you if you you should tell it hey work on this idea and and you sort of like forget about it for a weekend because you're just taking some time off, it's going to work on Friday, on Saturday and on Sunday. And when you come on Monday, it's like well in the past three days I did this, this, this. So it prompted itself to be like, okay, well where are we at? Okay, so I guess the website is still not done. So like let's continue building. What does it need left? Like well, I remember my owner said that he wanted to do those features. So you're like, let me build them. Okay, then it's the next day. Well, he hasn't responded to me. So let me do some SEO setup so that like the website is SEO ready, right? So and the user doesn't have to ask for that because policy is optimized to just make it to achieve the goal of the user. And so if it's like, oh, I want to build a company and get customers and it will set up everything for that. But if it's like, if in the future, it's like, oh, I just want to build this little personal project for myself, then it might not need to like do SEO anything. It's just about building the tool and maybe researching what are the best ways to make that tool better or maybe maybe the future could design it to different versions of design. And so that's the main difference is that it's autonomous. And that's and and by the way, autonomy is extremely expensive. It's also why other companies are not fully going in because it's expensive and it's scary. It's expensive because agents are expensive. Like they need to run for sometimes five minutes, 10 minutes, 20 minutes and just every turn this the cost money. And number two, there can be scary because they take decisions. And so it's like people are like, what if it makes a mistake? Well, I mean, it's better than like giving up, right? So it's like, if you if you actually are for sure want to build this thing and you're not going to give up and you know what you're doing, just use code code, right? And by the way, code code is generating like, you know, tens of billions of dollars here for and traffic. So clearly people love it. And I use it also. But it's some people that are like, I don't have any terminal like I don't know where to start like and also I get on have time. And then well, Paul, yeah, just it just works for you. It's fine. You don't have to worry about it. It's just going to work for you. As you do a better job in the clock code, it's not necessarily a better job. It just does it just works. It will it will figure it out. It will try, right? Which is hugely important for it sounds like your I see P you just at least at the moment a little bit more on the solo printer side. Someone who's got an idea who wants to launch it. And I mean, you've got over over 8600 companies active on policy already. So what are some examples of companies if you wouldn't mind sharing some ideas that are out there that people. Yeah, I mean, yeah, I mean, I mean, I was just like meeting with a customer. He has like a like a voice AI product is building right where it's like you can you can call with a phone number and then like it's like an AI sort of receptionist is building. I have like other customers that are building. There's a lot of AI SaaS products that are being built and especially the one generic money whenever I look it's often like AI SaaS products like you know AI ads or AI you know, yeah, for this, yeah, for that. I also have customers. I get one of a big customer. I mean, it's a big customer. It's customers that's been in the platform for a while and they really have power users is using it for his existing business. So actually he is managing a network of is a market space between people who need laundry services and laundry providers. And he was using a third party SaaS for that. I was crossing him like 200 bucks a month or something. And so he was like, well, he told policy. I can you just rebuild it for me. And policy, I just rebuild that solution for him. And then now he can add the features and make it perfect. He actually connected his own strap account because he doesn't want to pay the fee of that at 2 a.m. charging, which I think I'm going to drop actually. And yeah, and running his business on policy. So you know, it's like there's all these different ideas. It's also people running personal projects. So it's pretty varied. It's very varied actually. They're cool. Super exciting. And I mean, you mentioned pricing. So why don't we dig into the pricing curious how that fits in because it's quite a unique model. You know, I believe $49 a month base plus there's about a 20% revenue share, which is quite different than a typical SaaS pricing model. It's almost similar to more of like IC stake in companies or true partner in their building. So talk us through the pricing model and how you kind of lend it on that. And it sounds like you're iterating upon it too. I mean, it's very interesting because it's like, so where I landed on 49 was because AI is so expensive. So when I was building POSIA at the time, I landed on like, it was between $1.50 per task, right? So I was like, okay, I think I can bring it to a dollar. And so every night it has one task. That's $30 of AI costs plus like, are you giving them a West server for like, you know, five to ten dollars in database that cost me every month plus I give them some API keys because if they use OpenAI or like, you know, voice models, it should be embedded so they don't have to think about it. So I was like, okay, I can arrive at 50 bucks. So like, let me charge 50 bucks and make money when they make money. So that's the 20% right? So I thought that was like a pretty fair thing. It's like you pay for I sort of break even and I think over time AI costs going to go down and I can like make some margin there, but then I'll make money on 20%. So that's sort of the way the way I thought about it. And then if you want the AI to do more tasks during the day, you can buy tasks and it's around like a dollar to two dollars per task, which cost me a dollar to two dollars. And like, again, thinking over time, I can lower cost. Actually, the opposite happened like cost went up like the cost per task went up because just because, for example, like engineering agents working on vibe coded websites and could basis of customers getting bigger and bigger. Like it had to spend more and more time trying to fix bugs or figure out things up. And cost was sometimes 20 bucks, 30 bucks for off costs because then it would then it would it would go to like opus like the best AI models to try to fix the bug. And like, so the long story short is like, is that it actually even at 50 bucks, it's like it's not enough to cover the cost. So I actually started working with a so I'm working with like different companies that are all building infrastructure for agents in different ways. And one of the companies that work with Sapium, they sort of like, I'm I'm exclusively using them for a lot of like the APIs and the things I'm one of the biggest customer. And exchange is sort of build solutions for my problems because my problem is going to be everyone's problems when everyone builds those systems. And so one of the things we did together is like my debuels is is actually renting GPUs and doing our own AI models so that we get way cheaper more cheaper price. So that I could be, but then the price was like 100X less. So which is crazy. And when I realized that I was like, whoa, wait a second. So now it's a hundred. So now I was, you know, I was breaking even or losing money, but now I'm like, now it's a much cheaper cost for me. But then I was like, well, if my mish and then when I meet customers that paid a 50 bucks a month, they just love it so much. They're like, they're like, I had this idea in my head. And now I'm an entrepreneur and now I'm building and like, yeah, I've made a job, but like I'm I'm building and I can build different ideas. And you know, obviously, I'm a bit entrepreneur for a decade. And when I see someone becoming entrepreneur, you know, there's this connection and I'm like, yeah, I mean, now you understand. Now you're builder. It doesn't mean that your first company is going to be successful, but now you're builder and you're empowered. And and I was like, I need to give that feeling to more people, right? Because my mission is to empower the 99% like a billion people. I mean, this is a this is a big big goal, but like, this is the idea a bit. And I was like, well, I'm not going to get to a billion people if I charge 50 bucks a month because that's actually a lot. It's kind of scary. Younger people can really afford 50 bucks a month. And and so and so now my pricing model is like, I'm thinking, well, actually I have a free version so that people can experience just the autonomy. So they may not be able to like build all the things and like set up all this tribe stuff, but at least they can experience an AI building for you. So you give it an idea and it starts building slowly daily. And then and then change the pricing. We're like, you know, if you actually spend 50 dollars, maybe you can try five different companies. So you so there's less pressure for one idea to work and be like, well, I paid 50 bucks. I created one idea. They make money. It's like, just give it a few ideas, just a bit. So seeing what sticks, right? Because this is also what pulls you on power is that like you can give it a five ideas and it can build an MVP of all. And you can be like, well, this actually feels the best, right? I like that one the more the most. And on the 20%. It's interesting because like, you know, I'm thinking like if you know, obviously that was my way of making money. But then I'm like, okay, now that like I'm thinking way less on on AI, I could actually make a profit on on compute. Yeah, right? So I could see a world where now I can make money on compute. And I'm like, do I really want to take 20% of someone's business?
It's as aggressive and also, for example, the customer who's running his business on the policy, he uses his own stripe key because he was like, "I can't pay 20 percent." Like, "That's too much of my business. I don't have enough margin for that." And I was like, "Yeah, use your own stripe key to find, like just as policy, I'll connect you on." But I'm like, "I kind of want to be like Shopify, like charge like 2 percent or whatever, and just be more friendly and then maybe charge 20 percent in exchange for an investment." So now it would be like, because I was just talking to a customer earlier and I was like, "You love the product and he was hustling and he told me he tried to build the same idea like 3 years ago and he paid $10,000 to an agency and now he's rebuilding it and it opens his mind. Now, it finally worked and he was so impressed and now he's like, "I think about all its ideas." And I'm like, "Well, Polsia could identify this user because they could see that he's talking a lot, it's product works, he got his first customer, whatever, milestone." And talk to the customer and say, "Hey, I'll give you a thousand credits or whatever, like I give you an investment in like compute." In exchange, now I take 20 percent of your business. That's the 10 percent or whatever, right? So I get, I take that cut in exchange for actually I give you a lot of compute for you to really start building. Yeah. And I think that would be, instead of taking it by default, I would take it in exchange for the users agreeing to take an investment because that actually that user was telling me, he tried to raise money, he had friends and the VC said, "Well, you need a technical co-founder." And he was like, "Huh." So he doesn't have access to the capital markets. Like I do, right? So for him, he has to hustle on his own and I was like, "Well, it was very inspiring to talk to users like that." And he was like, "Oh, yeah, I mean, actually, they all put it like, let me invest." Like I want to invest in him, but to credits, right? Because I want him to use Bolsia and become a power user, become an ambassador. So. Yeah. But it's all evolving. You also learn from users, right? It's like, you have this idea in your head, you put it out there and then you talk to customers and you're like, "You know what? Maybe I'll tweak it. Maybe I'll make it different." Yeah, yeah, exactly, exactly. Especially in that early days. And I mean, it's such an interesting angle because it almost sounds like a wedge that you're using for angel investing and access to then trade off with compute. But compute does sound like it's your biggest kind of variable cost and almost biggest strategic dependency. Are you seeing even with these partnerships with infrastructure companies, like, "Are you seeing the path forward where you're able to stay profitable or be in like a net positive zone while paying for all this compute? Because compute is, as we know, quite costly." It is like the new currency. It's the new electricity. You know, I mean, it's actually straight up electricity at this point. I think what I realized is it's interesting that when I launched Borsia in December 2025 was the same month that Opus 4.5 came out. And when Opus 4.5 came out, I, like many others in the industry, were like, "Whoa, something changed." Like, the model is, there was a step function in the model's ability to reason extremely well on whatever you were saying. It's ability to use tools so fluently and just like being able to like not hallucinate and just like state of course. And so that's really a model that empowered, to me, to me that model was, is AGI, to be honest, like whatever definition we have because AGI is like this concept of like, it's like an intelligence that's broad, that I can do anything. And if you think about Opus 4.5, it can think about anything, it's reason about any concept you give it really well. And it can also use any tools. So it can actually act on anything, right? You can give it any tool, any MCP, it will figure it out. But fast forward three, four months, open source is always trailing through the six months. And now three of the six months open source is as good as 4.5. So open source to me is AGI and SOTA, like the 4.7 and GPT 5.5, they are AGI 1.2, 1.3, right? So it's like if you think about like what GPT 5.5 like solved like a math theorem that like was unsolved for like 50 years. And I think my toe, so whatever it's called like broke like the is used by the NSA to like figure out leaks in like security halls and like open source software. It's like my customers don't need that. They just want to create a little bit. They just have an idea on their head and they want to be create. They want to they want to create maybe a community. Whatever they want to build, they don't need AGI 2.5, you know? But they do need AGI 1.0, which is something that can actually execute and as well as a human. And so a month ago if you have asked me, I'd be like, yeah, it's tough to make money. It's going to be tough. Like I take the 20% I need to make sure. But now I'm like, you know what? Now with that with those open source models and then and that I can tweak to make it my model. I can actually fine tune. POSIA. I make it a POSIA model, right? That like is the POSIA personality and like and can be very efficient at like doing what what POSIA needs to do. I think that like creating my own model that is fine tune to be less expensive and also GPUs are going to get more and more efficient and models are going to get smaller and smaller. And so I think cost of intelligence will continue to drop actually. And so I do think that the future is it will be fine on margins. And I think right now it's more a race of like the like the best products, the best abstractions for different different people to embrace AI. I am trying my target demo is like non technical people. But if you go to like you know, there's other folks that are targeting the enterprise that are still not using AI that much or empowering employees in firms or employ, empowering lawyers and powering accountants. There's so many angles, but then the collectively with like it's about empowering folks to use AI to just output more and more productive and also be more fulfilled. And I think that's where I was, where is capital because then you need the capital to being able to sustain lower margins at the beginning while and also trying to offer things like premium that cost money, but it's also marketing because it's like if I can, if the haters on Twitter can all use the product for free and a tweet that would be like, you know what, it's actually pretty good. Versus right now they tried to be on body and like, well, I'm not paying 50 bucks and also like I don't like it, you know. So they can't really try the product until you experience it and you get the vibe of like, okay, well, that AI is actually, it's pretty cool. It's pretty cool that it's working, continue working on my idea when it was just, you know, yeah. I mean, everything you've shared, it's a pretty radical view, Ben. And you know, you're really betting on the price of compute and AI coming down in this kind of radical new world. If you're right, what does the new world look like? What does the future look like? I think that what I'm betting on is, you know, if you have a society where people can find jobs and they depend on welfare or like some new welfare that like is much higher because some folks are paying some, you know, government figures are way to tax like those conglomerate that are milking the economy or earning the economy. That's not great. It's a great outcome because, you know, people need to have purpose, people need to fill in the contribute. And I think that like, I think that as long as humans sell to humans, humans know what the humans want. Today, and by the way, in the future, humans will want different things. Because maybe there'll be more offline experiences, there'll be more, there'll be more, people want different things. Maybe the internet will become the dead internet, like this only AI. And people won't go anywhere anymore and they will find out the places to go or to hang out or to do things and to feel fulfilled. But what will always be true is that you'll need builders that build those new things that entertain others or save people's time or provide them value. And I think that like if you have a collective society that's educated that like, because what happens when you build a business is that you become educated on a lot of things because you're going to get an engineering into marketing, into finance, into operations, into support. You have customers, into business, into profits, margins, right? Investing. You know, all this concept become real because you have to learn them because you are trying to run a business. And I think if collectively your society is like, well, everyone gets educated business, then you get a much more, you know, you level the playing field in a way and then, and then you create a new economy where like actually people can sell to each other, right? You could see pulsias in economy, not just the tool, right? Where people can buy and sell businesses, people can buy and sell services inside of pulsia. You get, you know, I could see a world where like someone is like, well, you know, I want to build enough line thing actually. I want to build a bakery, right? It would be extremely hard for someone to, it's so much work you have to find a space, you have to do this is in that. Versus AI could be like, well, you know what, like, let me find, figure out for you, like, let me figure out like an affordable way to do it. Oh, let me hire stuff for you. Let me, you know, because AI could just find a staff, hire them, manage them, pay them, fire them if they don't do a good job, promote them if they do a good job, give them a rating. There's a way that AI could actually employ more people because, you know, if you told me today, I could love to open a restaurant, but like, I know it's about, it's not the best investment, it's the most gratifying thing. But never in the world I would spend the time to deal with all the obstacles to logistics.
But if I go tell you, hey, here's a 200K. I want to open like a little time. - Yeah, like restaurant with your-- - Super cute with my recipe. I mean, I'm a huge snob in terms of food, so I have to think about it. But probably something French Italian. Yeah. - Thank you. - But it's like, I would give that money 'cause I would be like, that would make me so happy. But I would want the AI to just handle all the operations and all the stuff. And I would just give it my taste. I would make sure the recipes are good. The ambience, the design. But the way, like the fact that we need to hire someone to clean, someone to cook, someone to open and close the shop, someone who we had to pay the rent and figure out the accounting, figure out the taxes, the LLC like, I don't know how to deal with that, right? Nobody wants to deal with that. And that's what the AI can be really good. And so the long story short is that I think the future can be very abundant and create a society that's more vibrant as long as like people equipped and indicated. And so that's why all this noise that make online is also about education because whether people like it or not, they now, well, what is this AI slop thing? Like, what is AI slop? Like, wait, what is this thing? And then they tried the onboarding and then for free I give them like an explosion of AI in their face, right? And they're like, that's crazy. I didn't know AI could do that 'cause I knew I knew chat GPT, but like you ask a question and it answers. This thing like just created a company for me. Is that a company? And then they can decide that I like it. I want to continue or I don't like it. And then, you know, definitely. And these kind of radical views been, I mean, first, it wasn't incredible and interesting future and vision that you have. And these kind of radical views, it takes somebody to actually execute on them just like you are. And to this point now, we're sitting at a where we're tentingly in an ARR. - I mean, inspiration came from like a decade of building to be honest. Like it's not, I think it's like the collective some of all my experiences. I think Paul Sia is like, is, you know, the way Paul Sia talks is the way I talk and the way I see entrepreneurship. Like the prompt is like really about the way I see. You know, so it's like it's a part of me. I think the design comes from gaming, like a game that I love, loved and love. And it's inspired by that and trying to make entrepreneurship more like a game. And I think it's, yeah, the inspiration, I mean, to me, like it's not like I can remember the day I had the idea was like I was in Montfouji. I was actually in Tokyo, like I was pretty stressed out and like about, you know, I was working really hard on my first idea and I went to Montfouji to like open my eyes and like mind. And I had the idea for Paul Sia. So this is how it happened. But it's also happened in a place where like I started being in flow and like not caring too much about what people cared. Like clearly I don't care that people think it's a bad name or whatever. And when you stop caring about what people care about and like start just building stuff that because it feels right and it feels, you know, that it feels right. Then I feel like the best, that's the best thing that happened. Right? Just meaning that customer and then hearing him having trouble raising money and then like just using Paul Sia to build things. And I was like, well, I want to be your investor because like, you know, I love that guy's vibe. And I was like, well, yeah, Paul Sia could be an investor. And it's like, and it just comes naturally. It's not like over thought. It's not like I'm not gonna make like a deck and like making a financial analysis if it makes a sound. And I'm like, I feel that that guy would love it. And I think, and I think it would be, and I think it makes sense. And let me try it, you know, that's the way I build now. And I think building in flow is the best way because you stop building for an outcome. You build because it feels right. - And often you hold on a great point there when you're listening to the customer, what feels right is what's valuable for the customer, aligned with yourself. So I love that. And you spent, I mean, four and a half years building cloud kitchens, you were employing number two after Travis Kalinick and Travis Kalinick was Ubers co-founder and CEO, an incredible builder. What are some lessons that you learned from Travis that you've carried forward to now building Paul Sia? - I mean, yeah, Travis, an incredible entrepreneur and was a big mentor and like a big part also of what empowers me to feel be so confident in building Paul Sia and grow it is because he has his incredible energy about seeing the future and like really believing in a certain future and like just going at it with a lot of boldness and a very inspired boldness, right? And so I think some of the lessons from working for him, I think first of all, when I started building like at this business line that I built under him at the Kitchens called Future Foods, which was like helping restaurants actually launch virtual brands, multiple brands online to help them bring more business to their kitchen, to their restaurant, which was really helping them during COVID, that was a very successful business. Because people were like struggling to get any money in and so products like mine would really help them. And I remember before, at one point I realized, okay, that product works and I think it can scale. And as Travis is like, hey, can I hire a salesperson to help me sell that product to more restaurants? And he was like, man, you don't want to scale a product that doesn't work, trust me. You are the salesperson, you go close 10 customers and then come back to me when you're close 10 and then if you're close 10, we can discuss. So a week later, I closed the 10 and then I came back to me and I was like, okay, cool, all right, now you can hire one salesperson. And I think that's scrappiness and then that humbleness of not trying to scale something before you're convinced that there's a customer really wants it. I think there was one valuable lesson. I think the second valuable lesson is like, champions hard, really cool champions hard, which is like, when you get punched down, you get back up. Like if you get back up every single time you get punched in the face and then you get pushed in the ground, you get back up, you're insuppable. And so it's like that resilience, you know, like for example, right now on Twitter people are hating and this and like that resilience you build of like, I'm gonna take the part that they say that is, that I think they're correct, but the rest I'm gonna just ignore. And I'm not gonna be, I'm only caring about any customers who's talking or like the people that have a point that really, you know, and I think that resilience is important to not be affected by the world, just be affected by the customer and then also what's right, because you know, the consensus can have a moral campus that can help you guide you to ambiguity, right? Because you know, for example Travis, when he was hit by the search pricing, there's the logic and then there's the ambiguity of like, how does it feel and what's right? And how to present it or the tipping, you know, no tips. People are like, well, you need to tip. It's like, another day you're selling to humans. So like make sure that like it's accepted by the human society, right? Or it's in, you know, you don't wanna be completely like, sort of like country, country. - Right, you speak there like that. - But yeah, these are some of the lessons and I think now in this phase of Bosia, it's like, it's about dreaming big, while staying humble, like while, you know, till looking at your step and taking one step at a time, but also dreaming big, which, you know, obviously looking at Travis work for five years or clearly like you get, you're like, okay, well, I guess, I guess it's possible to dream big. You have to be smart about it, but it's possible. And it just, I think it's the first step to think it's possible to then being able to actually accomplish it. The first step is believe it, yeah, right? And I think in parts, that's what Bosia is trying to do in this first phase, it's like showing everyday people that like, no, you can be like the guy at Tim Vellan Valley that building a startup. Like you can do, you can be that. Now, are you gonna be successful? A lot of them are not successful in Tim Vellan Valley. So it's not a, it's not a get-rich, quick scheme, but like just let me show you, I mean, Bosia, let Bosia show you that like you can be a builder and then you can decide what you wanna build or if you wanna build. - Yeah, great, great lessons. And one thing you mentioned was a scrappiness. I feel like you've taken it to a whole new level in a way, because you hit 10 million in revenue run, or 10 million in run rate. - With one employee, yourself. So why build just yourself? Because you've talked about, you know, thinking about pricing strategy, you've talked about taking customer calls, it's a lot. You've got investors now. Why not hire a couple of engineers or anyone else? - So first of all, like I'm working with four different companies that are building in the agent space. And they have engineers and they are building their own platforms that I'm using. And so I have a proxy where like, I'm working with them. We have a commercial agreement where like I bring them traffic on a specific slice of like the economy, the agent economy, and exchange the sort of like, provide engineers that are building solutions for me that I plug into Bosia. So that's sort of like a little hack to have help as I build, right? And obviously when you become a big customer, like people are really willing to work with you and help you build because you're a big customer. And so I'm a big customer on behalf of my customers, right? Which, you know, the more. So that's one thing where like, but of course, sometimes I'm thinking like, if I had my own engineers that like we're working just on Bosia, maybe this on that would happen faster. And like maybe if I had like a head,
ahead of this and ahead of that. And honestly, I'm thinking about it every day. (laughs) Should I break the rule and hire a full-time team? The reason why I'm staying for now, at least for now, is I think it's a great marketing story, which in the world, while marketing is so hard, someone on Twitter said it's performance art, right? It's like, I'm sure it's performance art. This is like a whole of a, it's just like, and I think that's not wrong. I think that's, I mean, it's not fake, but it is, it's me being good at marketing. - It tends to make sense. - It's like, it's like, you know what? If I'm doing, if I'm hitting those muscle alone, people are gonna be like, what the hell? How is that possible, right? If I start hiring people that, I'm just another dude who has a startup, right? If I stay alone, the story gets crazier, crazier. The question becomes, when does it hurt the company? Right now, it's like, we're like on the head, right? Because, but also on the other hand, on the other hand, me being alone means I need to leverage AI crazy. So for example, obviously, Polcii is all AI. Customer support is AI. Now, I made it very easy if you have an issue, and you talk to Polcii, it's like just on your behalf sends an email to support that can do more refunds. But if like support cannot fix your issue, it creates a bug report. And that bug report, if there's enough people talking about the same bug report, gets automatically solved by Polcii and then pushed to the code. And I doing that because I'm like, well, there's no other way. Like there's, because the people are tapping into too many edge cases, because it's offering to build something different. And they're all, and so I cannot fix all these edge cases at once. And so I need to do this. But then if I get it right, I'm unstoppable. Polcii is unstoppable, because then first of all, Polcii start building for the people. So it may take decisions over time that are just aligned with the paying customers. And the voice of the paying customers gets what it's built, which is what you want. Like you don't want like some random PM that just decides what's cool. It's like, just listen to your customers, right? They'll tell you what's important. So that's the net positive. And also I'm putting myself in the shoes of the customer when the customer has a bug and they call, and I, because I give them my phone number to all my customers because I meet them in process. And then they call me at midnight saying, "Vendus a bug." - Any all your customers in person? - I feel not all my customers, but I'm trying to meet more and more of my customers in person. Well, you know, if you go to my Twitter, I start doing, I film them with that content and then like with conversations and because it's good content to show showcase sort of like what's going on. But then I, then they have my phone number. And I tell them, if you have any issues, anytime, just call me. Because my email, I may have trouble seeing it, but if you text me, I'll see it. So now, once I, you don't have to say twice. - Yeah, that's total. - Then they will text you every single time's issue. So then I feel the pain with them, 'cause I'm like, "Okay, I have this podcast, but then I also need to fix a bug." So yeah, it's a phantrony, but we'll see how the flow goes. - Well, I was just saying to see how you continue to build it. And take a sense in your marketing mind a little bit. You've manufactured some incredible moments. Intentionally, they are very true and authentic, but you've really been intentional about how you've done it. And one of that was your fundraise process. Another, you've got the live website, which actually shows people building. Let's first talk about email, because I've talked to Pulsia via email, for example, Pulsia handles and can resolve most emails, but how much of your emails actually get handled by Pulsia versus when you have to step in? - Pulsia responds to all my emails. And then it doesn't really, I mean, for my personal, for support, if you go to
[email protected], then it can actually resolve. Anything related to customers of Pulsia, right? We funds, I'd more credit, fix bugs. But with someone emails me,
[email protected], I have another Pulsia that responds, but it's more like to answer questions with when I don't need to answer, when Pulsia knows the answer or when it's a common question. And it's also a way to buy me a self-time, because I get so many emails, I don't have time to respond right away, so at least the customer gets a response, or someone gets a response, right? But I didn't give it an ability to book a calendar and I might be hard because then, I don't have time to meet most people, to be honest. I always see mostly say no to, I mean, I just don't respond, because I don't have time. - Yeah, completely makes sense. And what were some other kind of moments you thought very intentionally about? I know the fundraiser's one that stands out for myself, because a lot of people have been skeptical of it. We've had some guests actually talk about agent to agent communication on the fundraising side. And say, in a corroborate kind of what you are doing, for example, on Hoffman of Flex Capital, said, the first meeting will be agent to agent without a doubt by the end of 2026. You were on your fundraiser's process, programmatically with agents with Pulsia. So curious to hear a little bit more about that decision and that process and when you came in and stepped in in person. - I mean, yeah, first of all, when I realized that Pulsia was working before, before it even moved to SF. And I could see that people really engaged, people loved it. I, all my friends started to use it, which was really rare for products like that. And people like, I really love it. Like I'm building this thing and building that thing, right? I was like, okay, and what had happened like in 2025, is that like I built this app that makes apps, like the blanks app. And there was a competitor, someone who built the same thing, called Wabi. And they raised like so much money. They raised like, I don't know, $100 million or something. Like crazy amount of money. And then I was thinking to myself, like, dude, you like, you build the same thing, but then you stayed on in your cave, like traveling alone and then you never really gave it a chance. Which means that like, you know, if they are successful, you're gonna be like, oh, I should have went to SF, raised more capital, be more bold. So then when I built Pulsia and I was like, okay, I really feel like this is a good idea this time. Actually, there's retention is much better. I was like, I don't want to repeat the Wabi incidents. We're like, they raised so much and then I didn't raise any. And then like, so then I was like, let me move to SF and I need to do, I think I may need to raise capital, but also I need to market the product. And I was like, well, if I want to market the product, I need to, I think since Pulsia is claim is to do so much, to run a company, which is crazy, what better way for the marketing to be about AI doing crazy stuff that usually only reserved for humans. And so the first marketing, I was like, well, if I make AI raise my round, that would turn heads. And so, and then actually at the time, I already, because I raised the preceded, I raised a million dollars preceded in the summer that I haven't really spent because I was just allowed. So I still have a million dollars in the bank account. And so I was like, I don't really need it. I don't really raise desperately. But I was like, you need to do SF and I need to launch that stunt because I'm like, even if I don't raise, I think it could make people talk. And so that's, that was the first intent. And what ended up happening is, yeah, so I said I plugged Pulsia on my email address and I was like, you respond to every email. We built together the live dashboard, which is supposed to be the data room for investors to be able to like see the data, 'cause you know, investor like, "Hey, show me your data room, show me your data, show me your retention." And so there was this live dashboard and then that showed the actual number, number of customers, number of the stats. And then the chat where you can ask questions and the answers questions about retention, about this and that. And I started tweeting about it and sent a few emails to a few investors about it. And then people started like emailing me, 'cause the investors did their job to just email founders. And so Pulsia started having conversations with all of them. And the more this was happening, the more people started talking about it, the more that live dashboard curve was going up. And so I kept tweeting about it and there was like a self-fulfilling prophecy because the more I tweeted about it, the more the numbers went up, the more people will email me about investing, the more like, and Pulsia was telling them that it was raising at a billion dollar valuation. And so people were freaking out and telling other people. And so it was a really great marketing moment. And there was so many investors emailing that like I couldn't respond to any of them. So I made a few in person. But while I was doing so, I was getting a flugged of emails. And yeah, and then I think people were getting convinced at that point, they were just like, okay, like we're interested, right? And so then of course I had to, I mean, then I answered a few of them and saying, okay, let's be in person. And then the deal went pretty fast because the company was so hot that like people were like, you know, they just wanted to invest. It's like after many in person, they were like, and also it's like mostly investors that like knew my investor. So like there was like, you know, there was this sort of like, 'cause then the day investors are humans. So they also need some validation about, you know, the background of it made some background checks on me and stuff like that. - Yeah, love it. Well, I think it's a master class in building distribution early and building it.
such a powerful way in getting a conversation and talk about you early. You know, what we see is a lot of the time now, distribution isn't an afterthought to product market fit. It's really how you get to product market fit and how you maintain it now since it is more of an ongoing state, rather than an end state. And so you really built distribution of the gate, which is what we're seeing now in the market of what needs to happen is building distribution, focusing on your go-to-market, even if it's one person, distribution is your go-to-market earlier. So very excited for you in the journey ahead. - Thank you. - Appreciate all the time, Ben. This has been fantastic. Thank you and good luck with this mixed state and journey ahead. - Thank you so much.
PodcastTranscript AI