Silicon Valley Girl: AI, Tech and Career Growth - $13B Lovable Co-Founder: Build a Profitable AI Product From Scratch in 30 Minutes

Episode Date: September 25, 2026

Anton Osika runs Lovable, one of the fastest-growing AI companies in the world, and he's done with the words "vibe coding." Vibes work for a demo, but people now run real businesses on Lovable. In this episode we go step by step: how to turn a tool you built for yourself into a real business, how to set up a backend without knowing how to code, and how to price your product. And as building gets easier every month, Anton names the one skill that matters most for an AI founder right now.*The website Anton improved for Marina:* https://marinamogilko.co/📩 Follow my Newsletter: https://siliconvalleygirl.beehiiv.com/p/the-one-prompt-you-should-run-every-month-d907?utm_source=spotify&utm_medium=description&utm_campaign=futureproof-sub&utm_content=one-prompt&utm_term=anton-osika𝕏 : ⁠⁠⁠⁠⁠https://x.com/siliconvalleymm⁠⁠⁠⁠⁠🔗 Instagram: ⁠⁠⁠⁠⁠⁠https://www.instagram.com/siliconvalleygirl/⁠⁠⁠⁠⁠💼 LinkedIn: ⁠⁠⁠⁠⁠https://www.linkedin.com/in/marinamogilko⁠⁠⁠⁠⁠📌 My Companies & Products: ⁠⁠⁠⁠⁠https://Marinamogilko.co

You're abandoning the word vibe coding. Yeah, Vibe works for a demo, but we're way past that. You don't build your company on vibes. This is Anton Ossica. He launched Loveable, where anyone can build an app without code. Less than two years later, the company is worth $13 billion. Together, apps built on Lovable now get nearly a billion visits every month. And more than a third of the people building businesses on the platform are already making money.

Like, I've seen those dogs build on Lovable. $300,000 a month? Yes. Why? How is that a business idea? I think the creativity is the biggest differentiator as the execution becomes this year. What everyone in the knowledge worker space should be building for yourself to make their lives easier. Everyone should build an agent for themselves. Is it okay if I do the prompt here?

Please, yeah. I want to see your prompt. When I tried vibe coding, there were so many errors even a year ago. Like this QA was all manual, you would run into bugs. Now you said you're abandoning the work. vibe coding. Yeah, vibe works for demo, but we're way past that. You don't build your company on vibes. And people are running their businesses on our platform.

I have these amazing founders. Sabrine, she's building a startup called Plink. And she wanted to solve a real problem that she saw in the world, which is that can I trust this person I'm dating? And she's been able to reach tens of thousands of dollars in revenue on something that's just for social good. And the beautiful thing that I think it comes back to is that it used to be so hard to build the software

and it used to be a lot of overhead to run your business and you can now simplify a lot of those things. And that's not something you do on vibes, it's something you do on a serious platform. And when you think about a problem that you're solving, do you have a framework in your head? How do you solve a problem that seems kind of big? Do you separate it into tasks?

Because, okay, like I want to build, this content machine, for example. I don't know. When you have a problem that you want to automate, what is your thinking process of how to do a step by step and whether some of the problems are worth automating and some not?

I think it's usually good to just do things manually and then be systematic about, okay, how can I break this down into pieces that I can test and see if I can get value from if something is broken, that replicates the thing I used to do manually. If you're over-engineering or, you're overly ambitious from the get-go,

then you might end up being very close, but there's some small part somewhere that doesn't work in the way that you thought it would work, or your solution wasn't the right one, and you want to avoid that. You want to iterate towards perfection. So do you think, because doors are becoming entrepreneurs, do you think people who are day-to-day working on some boring tasks

are becoming the best position for the next three years to become entrepreneurs? We're seeing from our customers, customer data that most people on our platform, they have 11 years of professional experience. And they've been, from that, they've become very close to a problem, that they've been frustrated that they can't solve with the skills that they had at the time. And now the person who's closest to the problem, they can actually go out and solve that problem. And that's what is one of the biggest unlocks of these new technologies.

Do you see that group of people generating the most money was lovable? For most of our customers, it's people who are building and they're already generating revenue themselves or they're tracking towards and they're building a business towards generating revenue. Most of them have experience. There's also a lot of young people who are building companies, as you might know, in this day and age. And that also often is connected with having advantages of seeing things with fresh eyes for the first time if you're a younger person. But we're seeing people across the spectrum, across the entire world, every continent and every country, building very serious things on the platform. Love it. If someone's watching and they're like, okay, this sounds incredible, I want to build something, can you give me some ideas?

What do you think of the top three opportunities to build these days? I think what you want to do is you want to solve a problem that is not just on the internet, but you usually want to build something connected. to the real world, a combination of a service and the product. And those two things together can actually make it really good. This is the type of direction that I would be going if I was starting a business myself this day. So it's even something that hasn't existed. I don't think you need to start something that is completely new either.

You can take a fresh take on something. You can go out and talk to people who aren't yet adopting a solution that exists somewhere and see how can I make it so exciting for them that they want to adopt it. It could be something in how they get their groceries or how they get things in their home taking care of. And you can personalize it for them if you're an entrepreneur with that hustler mentality. Finding a good idea is the first step. The next one is growing it into something bigger. I love learning how successful founders think and steal a lot of it for my own work.

The best cases go into my newsletter. It's called FutureProve. so you can repeat them yourself. In my recent issue, I show how the top founders use AI for everyday tasks. There are rules and prompts step by step. Plus a bonus. The one prompt I run every single month that helps me with my productivity. Get it right now? The link is down below.

Do you have any ideas what everyone in the knowledge worker space should be building for yourself to make their lives easier? I think everyone should build an agent for themselves. And if you do it from scratch or you use AI to build it for you from scratch, then you get a better understanding for how does an agent work? Because you can ask, how did you implement building this agent? So I think that's something we could build today. I also wanted to actually make a refresh on your personal website, if you wouldn't mind. Oh, please. Let's do it.

Absolutely. There was a specific challenge I heard about, which is how you handle everyone who wants to talk to you. and that send you emails. Yeah, yeah, yeah. And building an app specifically that helps sort that out with help of AI. By the way, let me ask quickly, when do you build an agent? When do you build an app? I think everything that needs data is an app.

And then if you want to oversee what is in the data and have some pieces that visualizes what this app does, then you, of course, want a graphical user interface. A recent thing we've added to Loveable is that for every single app, you can connect that to any agent harness. It's called Agent Integrations. And then it's almost like the data that you're talking to is a part of the agent. You can use Loveable as that agent. You can use Clowal, Chat, whatever. So these things are actually, I would say, merging, depending on how you like to think about it.

So, Marina, I said I really wanted to create a new version of your website. Yeah, and you know what I wanted to just, I don't know if LaVable is going to do it automatically. It looks the way it looks. It used to be much more stylish, but then we redid it for AEO, GEO, whatever it is. And we didn't really pay attention to design. We just paid attention to being visible by chatbots, yeah, and agents. So that's why it looks like that.

There is a primitive in lovable to optimize for AEO and SEO. Do I need to say it to lovable? We'll first make it look good and then we'll optimize it for AEO, okay? All right. Okay, so tell me something about what you care about with your website. in terms of the brand and the aesthetics. Absolutely. So Silicon Valley Girl is a podcast that helps people see this bridge between them and Silicon Valley.

Because for a lot of people, Silicon Valley is where where stuff happens, AI, whatever it is. My goal is to translate what that AI means for their daily lives. For an entrepreneur, that means building things faster. For a mom, that means managing household with some agents. and maybe in 10 years having a robot that does the laundry. For someone who's an artist, that means just managing the whole production flow with AI. And I interview founders and scientists who are building those tools and ask them to show me what I can do today to feel the magic of AI.

Great. Let's see if we can make a new version of the website based on that. Just so that it has something to reference, I think we can put in the URL to this. the current website. Now what Loveable is going to do is that it's going to research your current website. Look at how it looks, look at how it's implemented. And then it might, depending on how you want to use it, ask you some clarifying questions. Or it will just give you the first new version that we can look at and iterate from there if we don't think it's perfect yet.

Can I see the prompting that's going on there? Yeah, you can click and see what's happening under the hood. Here are the details of how it's thinking through finding the current website that you have today to be able to compare with that one. So the color grid, we have white, red, black. And I was thinking, we can go back to your original problem statement, build that, and then we get back to see how this new website is like that. Oh, the email?

There are so many different types of queries that come to our email. So my initial thought was, like an overview was going on, categorized AI intelligence to tell me, oh, this was actually major. Is it okay if I do the prompt here? Please, yeah. I want to see your problem. Okay, so let me tell you what I heard you say, and I'll have lovable listening. So you want a lead capture form for Silicon Valley Girl, and anyone should be able to reach out to you and brought different types of opportunities.

And then you and your team, they should be able to try it and see all the opportunities sorted by how relevant an AI thinks these are with a proposed answer. and then when they are maybe choose what the AI is proposed or they do some edits, they can send back the answer to the person who reached out. Is that correct? Yeah. The question is, should it be inside the email? Because we already have our email listed on several websites. Or do you think we should now just list the website so that they fill in the form?

Yeah, you could take care of the in-viled emails as well. That's a great question. Let's start with this. And then we can use the same flow and connect and read the emails as well later on. So we can do both. Okay. What kind of opportunities should people be able to submit? The thing is, so when I'm thinking about the forums on the website, I want it to be, first of all, as free flowing as possible.

To give people some freedom. Just write our own. I think in the interest of time, let's tell Lovol, be as fast as possible at implementing this as well. Shall I just write it down? That's just possible. And don't limit to any use case, right? Yeah. So we'll let it build.

Yeah, now we can just let it build. It's going to take a few minutes. You're like me. You're like, what's happening here? Oh, yeah, it copied my style really well. Lovable keeps asking you which design you prefer. And you can zoom in on them if you want to or just pick the one you prefer. I think I like this one a lot.

Let's do it. Yeah. So this, like a knowledge worker can just build for themselves. What separates something like this that is very personal from a business that can be built on lovable? How do you distinguish what idea is going to make you money? Like, I've seen those dogs build unlovable. I could not believe when my team was like, have you seen this website? The paintings you order with the dog?

I mean, $300,000 a month? Yes. Why? On dog paintings. Explain it to me. How is that a business idea? So the founder, he's obviously very curious and creative, and when he sets his mind to something, he goes through with it. I think that's the main differentiator if you can build a serious business on this platform or if you can just build a tool for yourself.

Both are good outcomes. But for building that business, then it usually comes down to how hard are you going to work to make this business work? and are you going to be able to make people who would be your customers care about you and understand that this is something exciting that they want to be a part of? So when I'm looking at a lot of businesses as an entrepreneur myself, I'm like, okay, even when I have ideas, I tell myself, especially if it's like a dog-like idea, like, oh, I'll make a dog look like a painting. This can be done with Chad GPT with an easy prompt, right?

How do I tell myself as an entrepreneur to keep pushing? Like, I don't know if you talk to that guy because he's, I think, the best example of it because I tried and I could do the same similar thing with Chad GPT. How do you not stop when you know Chad GPT, Claude can do this with unprompt? Yeah, so Damien, who built this

surrealism world website, he doesn't just create the images for the dog. He sends it home to you as a portrait. And I think entrepreneurship is usually about combining those new discoveries into a service or a solution that hasn't been built before. That's when you can really make people care about it. And that's what he did, combining the website, the payments, and somehow finding marketing the service, of course. And this is amazing because he added this extra feature and turns out there is a market for it.

So how do you, first of all, how do you know there is a market for it? How do you price it right? Even when it comes to lovable, right? When it comes to pricing, how do you think about it? And there are different approaches. You can be just your home in your garage. But the most reliable and I think fastest way to learn about your customers, which is key, is to get in front of 10 customers, of potential customers.

And tell them about what you're thinking of building and actually dig into what they care about. Just ask a lot of questions from your customers. Like what's a pain that you're experiencing, what their interest that you're, excited about. And then going in front of 10 customers and try to make one of them actually buy and pay for what you're building. Is that what you do as well? Do you still talk

to 10 people before I build a new feature? That's what we try to do. Yes. That's what the product tries to do. And you think 10 is a good number. I think 10 is a good starting point, yes. And then in terms of pricing, I think you should get 10 really big fans of your product.

And one way to know that they are really big fans, is to not make it too cheap and not make it free because it's only if they actually pay that you know that this is something they really care about. You introduce that friction to know that they really like the product. And then you should hope that they spread the world, make them feel special,

and make them spread the world about the spread the word about the product that they are actually paying their hard-earned money for buying and then hopefully you get more customers from them. And over time, figure out how to scale that distribution. Yeah, and also knowing how big it can get, because if you have a lot of ambition, sometimes you think, oh, maybe this is a too small problem to solve? Or do you think a good entrepreneur starts with a very small problem and then finds a will to scale it? I think it comes from both sides. Some of the most authentic products, I think they start with

a problem for yourself. I was to some extent passionate about building lovable because I I felt that there's so many ideas, both from me and people around me, that will never come to fruition because of how hard it used to be to build software. And that's what we started, and being able to give anyone who's really close to the problems and who have that ambition and ideas, the new tools to do it, that's what our mission remains to this date. Okay. Let's see what it's building for us.

This is building, work with Marina Brand Deals, anything. Tell us your name. What's it about budget or time? I'm lying. I like it. We can polish the UI later, but just from your point of view, the first reaction, what do you think about this? I think it's great. Cool. I like that it's also opportunity, so it's like filters for serious requests.

Let's test it out with some example data. Wow. Only that. Only such opportunities. Okay, what am I proposing? Do you want to sponsor the podcast? Sponsor, yes. Sponsor pod. Okay. Send opportunity. And then here. Here we can see the inbox now, right? Yes, great. Well, I'd love to see the budget right here. So I can prioritize.

Okay. Let me show you how you can iterate a bit on the visuals by creating a new draft. So what you would do then is you go up and you click new draft at the top and the title. Here, right? Correct. Oh, so it's going to create the same version but looking different? Yeah, and then we can choose which one we prefer. And then we can use the microphone again and say, hey, let's say,

let's improve the styling a bit. You can maybe find Marina's Beehive website and take the branding from there, which I think looks pretty good. And let's say, while we're iterating on the branding, we should put your face on there.

So if you just search for your name, taking a copy of photo, we can paste it back into lovable, and then we can ask you to put on your face in a way that makes it feel more personal. Great photo. To the son, yeah, signature.

And we just put it here in. You can just drag it in from up there, I think. Add it to branding? Yeah, add it at the top. To the top? I wonder, like, will it look stylish enough? I think it will understand that this is a photo of your face. And we will add it to the top.

Photo was added. I think it looks better. Do you want to check if we also want to add the new branding, if we like that one, by going back to the other draft? Ah, this is so much better We can also look at the website Meanwhile Do you want to look at it?

I like that you're also multitasking Yeah Okay, so we asked for a new take on your website What do you think? I think it's great I think it looks amazing Question, is it GEO and AO. That's a great question

Click on SEO and AI search And here you can do a bunch of things The thing that I think is This simplest thing to do is to just tell Lovol, hey, we want this geo optimized by doing a scan on how can we improve discoverability for AI agents and Google. So these are example questions that you could do if you want to research. How can we differentiate more for a specific audience?

But now Lovol is going to run your full SEO and AEO. Can we make sure that the team receives an email when a request comes in and that people, who submit the form get an email that it was submitted. I think it's fun with all this listening. I know, I have too many assists. But it's whisper flow and granola. Yeah. Same as me.

Yeah, because Granola is recording. It's going to give me feedback on how I did on this podcast right after I have an agent that grabs the transcript and then gives me feedback. I should have this. And whisper flow is just. In my meetings with Senka, she can roast me. Yeah, actually it's so much. Because AI is more empathetic than my brain, for example, to me.

Empathetic? Empathetic? Yeah, it's like, Marina, you did well. It's okay. It's so nice to me. Yeah, you'll mix this up, but it's fine. Next time you're going to remember, it just makes me feel better. Because if it's me, I'm just roasting myself.

Okay. And here you can see some more of the primitives that is making lovable into a platform for you to run your company on. So you, obviously, assume you can store data in there. you can handle emails, but you can also do things like payments, connecting to your existing emails or connecting to any other tools that you might use. You might, for example, want to get this posted in Slack when there's a new email coming in. Yeah, because now we have, now that I'm thinking, we have everything that's running, like inside different AIs and projects.

If we connect all the emails, all the corporate emails to this and make it a hub where you see all the brand deals and what's working. Like if we have email for producers that are working on a particular video and there is a brand deal, like it all gets connected. Interesting. So it's like building a back end for your business. Yes. And this connective layer between the different things as well.

How much more expensive is this versus using just a cloud project for running your business? Because like in terms of tokens? For the building piece, it's very similar in terms of costs. And what you get as well, which is when we included for free, is the hosting for the projects that you're using, which is very scalable on the hosting side. And you get these things like the analytics. So you can use level for free every day with a certain amount of credits. And then if you want to upgrade, the default is $25 per month. And how complicated can this setup be when I'm building something?

Can you draw me a line where it stops being something? that can be done unlobable and I have to hire a developer. Does it exist? I think so. That line. The line is probably moving, right? Yeah, the line is very much moving and it depends on how dedicated you are and how much of a creative problem solving person you are.

If you're dedicated and curious and really want to learn these things, then the line is very far out. For example, building Facebook requires you to have a lot of load, a lot of hundreds of thousands of customers at the same time and do it globally, that is outside of the line. But most of the software that I'm interacting with day-to-day, you can build them in lovable.

What about something that I built 12 years ago with the help of different developers, study broad booking platform? I think that's a great example. You can build it with lovable? And the thing which I hear from our customers is that they are usually the ones who understand their users and their customers are best.

So they're closest to the problem. And that's why you often end up with a very good polished, very well thought out product compared to if you have very slow cycles where you have to hand off every few weeks. I remember that. So that's a big part of the advantage on building on this platform. And what we work really hard for is that it just works. And you can focus on your customers,

the product vision that you want to have. And because ideas can be deployed that fast now, do you think ideas still matter these days? Because remember how people used to say, idea doesn't matter because you have to build it, you have to deliver on it. The ideas are becoming more important. I think the creativity is the biggest differentiator

as the execution becomes easier. And you can even, if you're curious and smart about it, You can figure out how do I make lovable run in the background and do work for me when I'm not there so that you can really focus on the next big idea for your business and lean into the creative parts. How do you develop your creativity? Any ideas? I love that question. Do you go to museums?

Look at art. Being bored, I know it's good for creativity. But I think every person has creativity within them. and the best way to improve, increasing creativity is, you know, sophisticating it by trying to build and by taking your ideas and putting them to the test. And when you're doing that feedback loop,

that's actually when the creativity that everyone has becomes more refined and more applicable and you actually have more useful ideas. And when you have so many ideas, how do you, like when do you stop and tell yourself, this is what I'm going to work on, just because in this world, where it's easy to build ideas, and you have so many ideas, how do you select the one that is really worth pursuing? I still struggle with this.

Well, obviously you build this. So sometimes you want to really push through, even though you realize that the problem was harder than you thought. And sometimes you should say, hold on with that idea. I'm going to take the next idea, and then I'm going to take the next idea until you find something that really works. The way to do it, I think it's just a learned skill to develop in terms of, is this thing, how big can this thing become?

And why isn't it working as I wanted to work yet? Is it just me being out talking to more people? Is it me being better, have to be better at marketing? Or is it that the product has some bugs and glitches that make people not retain? I think it's being good at spotting that bottleneck. Why is the idea not working yet? that makes you become good at deciding, should I continue? Should I take the next idea?

Yeah, and I also think the scale, because I read your story how you realized that you're going to build the last piece of software. I think when you have this major idea that's so transformative, it gives you more energy to try and pursue it, right? I think it's very important to lean into what gives you energy. And to me, that is to a large extent the people I surround myself with. and are there people that are giving you energy that you can work with and then you can usually run through walls together to completely different extent. Totally. How do you think about competition?

Do you ever have these thoughts like Chagipiti, Claude can build apps? I don't think so much about competition. The platform we're building is becoming this perfect place to run your business. And we're not doing code generation. We're doing this creation of businesses for our customers. And it's still extremely early days in this complete agent development. So we're not actually going to see a lot of companies like us continue to grow at this extremely rapid pace. And what do you tell yourself?

Yeah, when it's so competitive. No, what we focus on is to look at what our customers are trying to do and how can we really make the best possible product and platform for them to reach their objectives, which is to build a business and then to run. and operate that business and actually help them with growing your business as well. And I showed you there that you can make your website AEO optimized. That's one of the pieces that makes it so that there's millions of visits to lovable apps. I think there's actually a billion visits every month to lovable applications.

Yeah. Do you have any fears at all that when AGI comes? I don't know, models start building for themselves. like when you think about five to 10 years ahead. Yeah. I think we're going to be in a world where focusing on your career and making money is going to be less important to many people. And it's more going to be focused on how am I a good person to those around me and to my community. But trying to build a business is still going to be one of the key ways of developing yourself,

learn skill set and creating valuable things for those around you. And what we're working towards is that the overhead and the costs of making a business run is going to be so low that there can be much, much more companies that exist. All right. Well, let's see what is doing for my business now. And it's like a CRM for managing those, right? What's the 10? What's the 5? We asked in the beginning for scoring them according to AI.

Before we wrap this up, I have an idea in how we make sure it actually works. I would just tell Loveable, can you test this end to end? end, both submission and answering the email in the inbox. And then you can tell me what we need to check in our inboxes to verify that it worked. And if you run into any problems. We're doing the end-to-end test. I just want to queue up and tell you something that I think is very important. And it's that we do a deep security scan to make sure this is ready to be published to the internet.

Do you want to just ask for that? Do deep security scan? Yeah. And I just send it. Okay. We asked it to do an end-to-end test. It submitted a fake pitch. It tells me to check my inboxes.

And it didn't find any errors. So it does like QA. It is some QA here. And it said it has sent two emails. Brand deal sponsorship. Beautiful. And then. Did it reply to the email as well?

It did. Okay. So here it actually decided to say yes to a collaboration. Nice. But you can like ideally I just see before. sending. But that's great. Marina, as you can say, I love building things. Yeah. And I hope this is valuable to you. Totally. It's been fun to build together. Thank you so much. Thank you. Let's wrap up with a step-by-step plan for someone who's launching a business with Lovable. So we talked about figuring out the idea.

We did the whole demo coding or creating a platform. What are the next steps that they should be taking? The first thing is to have an idea. And I think you get that by taking something from your own life or from people you're close to that experience a problem. And then you find your idea for how to solve that problem. And you try to find a few people that are willing to pay for your solution. And when they're actually willing to pay, that's a great sign that they care and they're excited about what you're doing. But you still want to make sure that they keep coming back and they keep want to pay and it wasn't just done from excitement. And when you've gone through those steps, that's one example of clear product market fit. And you should be then starting to think,

what are the most effective ways for me to find more customers like this customer? And think about where are they? How do they find new products? And what do they care about? What is it that's unique about me, that I can tell us a story to make those people care about this product and start scaling into a real business that makes serious money? Which is mostly social media, I guess, this day. The answer would be to the marketing question. I think there's this fun Twitter about that everyone, a carpenter from a small business owner to tech companies, they need to be on TikTok and creating reels.

I don't know if that's true, but I think it's a fun meme. It is. Well, a lot of businesses are starting on social media. Thank you so much, Anton. That was very practical and hopefully we inspired a lot of people who, or at least some people, to start and build something. I hope a lot of more people take the leap and they go and build. Yeah, awesome.

Thank you so much.