Transcription
Everyone worries that AI will replace humans. Mistake. Sellers need a virtual employee who will make them cooler. Well, roughly speaking, your product is in the top for 3 weeks. The total revenue of our clients is over 6.5 billion. The service is purely for managers. The system sees all their flaws and mistakes and highlights them to management. Which of these business models has already become commercial? 17 payments of 10,000 rubles. What you've done is useless to anyone. What's the request? International development. How to become a unicorn company? $1,700 turnover. You don't have a funnel and you don't have traffic for your own product. Traffic is not a problem at all. In a company, besides a product spirit, there must be a commercial spirit. I've invested about 10 million already. No, for now, we need to close the business. The system works, it delivers results. This is a way to talk yourself and the team into it. Your most likely bet right now is, come on, Ranets, tell us who you are, what you do. I am 34 years old, I am an entrepreneur. My current business is the thirteenth in a row. Before that, there were both failed and successful business models. I am an oil engineer, I worked for 5 years in the oil and gas industry. I've worked over 50 professions during my student years, during my life. 50 professions, yes, around more even, probably, most likely. Well, student years, where they pay, there they work. What's the coolest profession? And we were assembling New Year's lighting. I was an installer, who installs these on buildings near St. Petersburg. Uh-huh. I assembled a team there. Uh, they paid me 150 rubles for them, and I told them 100 rubles, like, and the model was very cool, but not the coolest job. Climbing facades. No, I'm the foreman, I assembled them, the installers assembled them already, so it was fun. As for the current business, it is the most successful. We have two main directions. The first direction is consulting for commercial sales departments on marketplaces. The total revenue of all our clients who work with us for 2024 is over 6.5 billion. In 2023 it was 2.6, in 2025 we plan 11. And our second major direction, which we hope to discuss today, is IT. We have our own self-purchase service. I don't know if you know what that is for marketplaces. We have 15,000 bots, 3,000 live people, eight million-plus cities. We buy ourselves out in huge numbers on marketplaces. Well, roughly speaking, your product is in the top for 3 weeks. Our flagship product is Sirena Ai. I position it as a virtual employee that analyzes the current performance of a store and recommends to managers what needs to be done to increase the store's efficiency. Okay. What is your request today? What hurts? Those who were involved in IT, the product that was invented at the beginning and the product that exists now, now they turned out to be completely different products. My request is how to become a unicorn company in this ecosystem. You don't need to become a unicorn company now. You need to show that it works at all. Uh, the first step. Product Market Fit. I think Sirena. The most interesting product now is Sirena AI, right? Yes. Everything else cannot become a unicorn. Uh, obviously, these reviews will not become a unicorn. Uh, the fact that you help sellers, do consulting for them, that also will not become a unicorn. What can become a unicorn is Sirena, right? Accordingly, you have a hypothesis that sellers need a virtual employee who will make them cooler. Absolutely correct. It's a sales director who monitors all indicators and says what needs to be done. Do you correctly understand that all three companies grow from one root? Yes. Yes. Describe the essence, if possible, the core, like, what lies at the very foundation and how this branching occurred. At the core is the automation and digitalization of business processes related to marketplaces. I came into this business from marketing, so I understood that without digitalization, without end-to-end analytics, without process automation, you won't win. There is a human factor, there are a lot of changes that happen on marketplaces all the time, everything is different. Therefore, this allowed us to create our own tables, our own reports, our own bots at the initial stage. And all this was transferred to the Sirena infrastructure, which we have now. So there is a certain expertise in this area, which you are experimentally trying to build in three business models. Yes, absolutely correct. We have very serious cool cases where we x10 stores. There is X10, X5, stores grow exponentially. Is this a period of experiments now, meaning that at some point only one will remain or not? Well, what an interesting structure we have. Sirena 1.0 was just an algorithm. If this happens, if that happens, do this. In the second version, we connected agents. It already determines which parameters are more effective, which are less effective. The third version, which we are developing now, is an open LLM that can be suitable for any marketplace. You feed it data, you give it the final result. It determines based on historical data what influenced the increase in profit, the decrease in profit, and teaches managers to repeat it further. We initially made the first version of Sirena for sellers. Well, supposedly they will use it and everything else. We released the product to them in July, and they look at it: "Well, we'll give it to the managers, let them figure it out." They found the product unnecessary. We were like: "Okay." In November, we did a second pilot, the service was refined purely for managers. They like everything, it's cool. We conduct customer development. What was the social problem? They are worried that the system sees all their flaws and mistakes and highlights them to management, and they sabotage us. Okay. Well, look, let's still sketch out the general context so that the picture is more or less clear, comparable for us. So, there is certain expertise in the field of automation, artificial intelligence, and marketing, as I understand it, right? At its core, you are trying to branch it into three directions and make them business models. Has any of these business models become commercial? That is, is there any revenue, profit? What are the key indicators now? People in the team, commercial indicators, those you can disclose. And I was specifically looking at the report. We have 337 system launches, 97 gave us tokens, and 64 are used daily. 17 payments since November. 17 payments, yes? At 10,000. For 10,000 rubles. Yes. 10,000 rubles is the cost of the whole thing. Okay, I understand. And it probably costs like a cast-iron armored train. I've invested about 10 million rubles already. Yes. Well, look, at the moment, from what you've said, your problem is not how to build a unicorn. What you've done is useless to anyone because you tried to sell it. In one way, it was rejected because your hypothesis is that there are people in the world who think today it would be great to have an artificial intelligence employee who would help me do this business. Right? That's your hypothesis that someone is looking for virtual employees. What are the two more main ones that we covered when working with Sirena? Let's say, since we are users ourselves and beneficiaries, perhaps it's correct to say, and all our clients who use it. The growth rate is simply insane for everyone. Well, there are no mistakes in your business. Everything is digitized: advertising, traffic, SEO, reviews, prices, this system tracks and monitors everything on a daily basis. Accordingly, the product, in your opinion, works. Look, Sirena as a product. We can confirm that this product works. It gives good recommendations on what to do. Yes. Yes, 100%. But it can't do it itself yet. It can only say: "Do this, do this, do this." Yes. Yes. You know, I recall the story with online education. In the early days, when we started, it wasn't as fashionable as it is now, we walked around like flag-waving enthusiasts. So, imagine, you can learn anything, anytime, and from anywhere. And everyone agreed, like, yes, that's really cool, but we will never learn anything and in any way, because, well, learning is painful. It might be the same here, that it seems like, on the common sense level, it would be really cool to have artificial intelligence as an employee. And the more, the better, it's cheap. But on the other hand, you actually face the problem of implementation, both from the decision-makers, because it needs to be somehow integrated into the architecture of a company that already has a certain design. And from the perspective of, especially importantly, those who are displaced by it. It's like, for example, before ChatGPT, AI existed, but when they made the context menu, the user experience of using the product exploded. And our product, in my opinion, is at the same stage. We haven't tested large traffic yet, we haven't let a lot of audience in. These 300 people are just from my audience, from my business breakfasts, coming in on a small scale. We conduct customer development weekly and improve the product. I'll tell you something cool now. Why do I see development in this? I saw this from BMA. There are three main bases in business. I don't know, everyone probably knows this. You have a product or service that you provide to the client, be it sales on marketplaces, consulting, services, anything. You have some kind of funnel with some conversions, everyone has it differently. You have traffic that you receive, also different. And the task of business is to get cheap targeted traffic, put it into the right funnel, and get the required number of sales. So. And what does Sirena analyze? It, well, a priori, you have a good product in a normal quantity or a good service in a normal quantity. Then there is a funnel. The funnel has different stages. And each stage has, well, a drop, a drop, a rise, drops, CTRs. They depend on many factors. And you have traffic, where different types of traffic also give different types of funnels. Well, roughly speaking, organic traffic has its own funnel, advertising traffic has its own funnel, external traffic has its own funnel. And the task of a specialist, for example, like a marketer or a commercial director, probably, is to get targeted traffic into a stable funnel, get sales. If your funnel is stable, more traffic, more sales. And Sirena is engaged in analyzing what your funnel is, how to influence these parameters individually for each, to increase each CTR, and how traffic quality is processed with this, and other external factors, like reviews, market price, and everything. And at the output, we get sales. Simply, Sirena, I was collecting analytics, it is suitable not only for marketplaces. Well, it is suitable for all marketplaces absolutely. We can do it on Amazon, we have already analyzed them. We can do it for Shopee, for Lazada, for eBay. They all have APIs, they all have the same business processes. And everything will work the same way. But you can go even further. You can just leave the marketplace market, go into e-commerce, it's all the same: traffic, funnel, sales. And that's why we're making an LLM, not services, like agents are made now, a bunch of agents. Why are agents made? Because there wasn't a product on the market that met our requirements. We've now assembled five agents from crutches that serve this ecosystem. I understand. Please tell me, what is the request after all? I'm trying to understand. How can we help you? Well, incidentally, it would be desirable to make it interesting from the perspective of a case study. Guys, who don't want to go through life alone, I'm gathering small groups where we go on the most incredible trips. This year we're going to Nepal, which everyone should see before the end of their lives. Then we're going to the Great Migration in Africa, where we'll fly over herds of animals crossing rivers that crocodiles are eating. Of course, the main highlight at the end of August, beginning of September, I'm going to the North Pole with a small group. And in such places, in the most unforgettable experiences, we will think about the most important questions of life. Be sure to follow the link under this video and choose the program that suits you. Bros. In short, everyone worries that AI will replace humans. Society as a whole worries about this. But our architecture allows us to create not just agents, but avatars. Roughly speaking, you can digitize any specialist, expert, and they will rent out this avatar to other companies. We have this platform, well, like, it's fundamentally ready for this. We can rework it a bit. Any employee can go to a marketplace now, digitize their knowledge, and rent out versions to other companies. This can be done with any business absolutely. But now many companies are developing exactly these digital employee constructors, where you can assemble an employee for any business, train them quickly. It can be a real estate agent, it can be a hotel receptionist. These are constructors that are not specialized. Your product is still specialized in something specific. Now it's tailored for marketplaces, because I'm in the market myself, so it's easier for me to develop it here. What's the question? For example, should I focus on creating AI agents so that others rent our AI agents, our Sirena, and scale their stores? Or should I open the market so that the market itself starts creating its AI avatars and renting them out to the same companies? For example. The second question is the development of international business. I've even collected reports, if you're interested, you can use them yourself and see the number of marketplaces in the world now. Well, look, what I don't like about your request right now is that you have 17 sales, $1,700 turnover, and you're already thinking about expanding your business to make not only AI employees for marketplace sellers, but all employees and all countries. I think that's quite strange. And essentially, you've just implemented an IT system for your existing clients. You have those clients there, you've installed this, like, brain, and they've paid you something. This is not the basis of anything. Well, it seems to me the question is premature. Right now, in the world of artificial intelligence, millions of experiments are happening simultaneously. The whole world is busy with this. The whole world. Everyone is doing this. Everyone is trying to implement it. Here, everyone is looking for where it works at all. Here it doesn't work, here it works, here it's better. And so on. There are people who do one small function, for example, a hotel receptionist, and manage to make an economy out of it that it worked. You have a working product already in the third or fourth iteration, which you haven't actually sold. Look, you've drawn a cool diagram for us with traffic, funnel, and product, but you haven't made this diagram for yourself. You have a product, you don't have a funnel, and you don't have traffic for your own product. What have you done to sell it? You already have a working product. Why aren't 500 people, 1,000 people using it? Traffic is not a problem in the market at all. Now we have a very easy entry into the service. Anyone can scan a QR code now, and in 2 minutes you'll have Sirena working. Yes, the entry is super light. We can do it for free. Now, my team tells me: "We need to increase the client base, rather than sell immediately." Like, drive traffic and sell for free. Just like Vyr developed, 100,000 users use Vyr daily. It holds a quarter of the market. It works. Have you tried what you're describing? No, we haven't sold yet. Look, since the task of this conversation is for the guys to have an interesting dissection, like in an anatomical theater, forgive me for the comparison, and for you to get some benefit, then you'll allow us to be politically incorrect, right? To say some things that in ordinary life should be said with more courteous words, like "my dear sir." Excuse me for that. So, look. First, we've been talking for about 20 minutes. I still haven't fully understood what you do. I mean, I roughly understand that it's artificial intelligence, employees based on this, but I haven't formed a clear story about what it is, how it works, and why, from your point of view, it can be a business, although I'm trained to understand such things. So, the first conclusion: if the founder cannot tell about it well, then, of course, his sales team will not do it better. It will only make it worse. Then, the conclusion. Nothing strange about such poor sales, such poor implementation. Then, the hypothesis. Well, again, looking at you and comparing with my own experience, perhaps I am looking at another enthusiastic enthusiast, a productologist, as I call them, who saw some technology, likes it, and is trying to fit it in somewhere. It doesn't matter where, not thinking in business categories, thinking in terms of the interest of digging into hardware or software. The whole nuance is that the system works, it delivers results. The client who uses it is satisfied, and sales are growing. The system is effective. When did the first client appear? In July. In July. And since then, 17 connections. Correct me if I'm wrong, but so far, look, there are, if again from these frames, there are simple ways to determine the success of a startup. Well, one of them is 4T, right? Technology, Team, TAM, and Traction, right? Not in that order, but first, TAM, that is, market size. Second, technology that creates something substantial. Team, you haven't said anything about the team. We don't know at all how well you understand this. Perhaps you are generally random people in artificial intelligence, and then this is just not a conversation. And fourth, and most importantly, traction. Well, traction is still very modest, so we are trying to dig. What is the reason? No, for now, we need to close the business. You are testing the hypothesis, whether anyone needs an employee. It's not a fact that people everywhere in the world are sitting and saying: "How good it would be to have a robot that helps me." Yes. So, I can give you two cases. We had the first case with real estate companies. We created artificial intelligence for real estate companies that optimizes, all the same. We tried to implement this super commercial analyst in existing agencies. They reject it because they have different beliefs, they don't believe in artificial intelligence. And it was easier for us to create our own company that uses the latest technologies than to retrain those who are there. Simply create a more efficient company and displace those who are less efficient. The second case, the same with lawyers. We created an AI lawyer employee and sold it to law firms. And they all didn't buy it, even if they bought it, they didn't use it, they used it incorrectly, because companies change with difficulty, it's not so easy to change a company just like that, implement it and that's it. Accordingly, we have now created our own legal company, which uses all the latest technologies, which will simply displace all others due to its efficiency, right? Accordingly, perhaps one of the options, perhaps you will create your own sellers who are simply more efficient and will displace sellers who are less efficient. This is also a possible option if if they buy poorly, right? Because in Russia, how many? 400,000 sellers. Yes, like, how many do you need for this business to be good, tangible. That's 5,000, it's absolutely realistic, 1% of sellers, right? Provided that they need it and are ready for it. Well, without us it's worse than with us. Look, here's another manifesto, right, about whether AI employees will be actively used in the world? Of course, they will, but this is a manifesto, right, and the question is, will you be one of those who will benefit from this kind of use? A big question, because one more thing, I suddenly realized now, Oskar, we have diametrically different experiences in many ways. Interesting, right? That you always, you always repeated models, and I always went, well, I'm probably dumber. I always went in almost always new directions, right? Like online education. We spent 3 years inventing a business model, invented it, and then they started copying us. And here's an important point. What's important to remember is that in your inventiveness and in the search for new innovative businesses, you are not alone. There is something that I have called the pioneer dichotomy for myself. That is, as long as you are digging and trying, for example, to invent how to monetize this artificial intelligence employee, no one cares. Well, like, that is, someone else might be digging, someone died on the way, then you also lost motivation and died. No one cares. But as soon as it happens that you, fortunately, most often from the survivor bias, have put together a business model, as happened with us, say, with Ontologi, then with Foxford, a huge number of others immediately latch onto you. You, that is, you rise into the radar visibility zone, corporations appear, other startups appear, which have the energy, drive, and ability to easily reproduce. Well, Oskar comes and says, like: "Oh, great, now it's clear how it works, it's clear that there's a market, let's go." And this is the next story, right? Even if we dig something up now, we're trying to dig something up now, probably, then your main problem is, if you are successful, what will happen when you enter the radar visibility zone, when OpenAI sees it, when Anthropic sees it, and all these guys who have tens and hundreds of billions of dollars in reserves. I think the first ones to see it, look, the most fundamental risk for you is that the marketplaces themselves, Amazon, and Alibaba, and all of them are sitting there and thinking about what AI products they should make for their sellers. That's their job, actually, for a marketplace to have an environment where different players in different circumstances against their philosophy. They want everyone to be on equal terms, accordingly, with the same tools, with the same. Now you're like hackers, right? Like, you've done reviews, raised stars, a product card. Essentially, marketplaces are against this. They don't want you to fake reviews, make fake orders, and so on. This is essentially you hacking, you're fighting the system. And the biggest threat to you is not Antropic or OpenAI, it's the marketplaces. Every marketplace now has 100 engineers whose job is to make tools for sellers. 100%. The question is simply how well they will do it, right? How quickly, how well. Uh, perhaps you will do it faster, right? But look, this is the main risk. We don't know, we don't know if it will happen or not, you can try. Also, we don't know what artificial intelligence will have, what it won't have. Well, from the perspective of your question, whether the general constructor of all AI employees that can be made, I do not recommend doing that. I do not recommend it. In my opinion, your most likely bet right now is specialization, very narrow specialization, because for these general employees, these people with hundreds of billions are coming, who train LLMs for huge money, right, who will do big things. Super specialization. Now I see, look, either this business model doesn't work, people don't need AI employees in this sphere, or you haven't tried hard enough to sell. You haven't tried selling at all. Mistake. This is strange. This is a huge mistake. A huge mistake. Essentially, we should immediately remove this from the stage. And until you make attempts, thousands of attempts to sell, specifically a pitch. You as a founder, pitch to clients, not investors, but clients. I pitched 100 times, 5 gave APIs, 2 bought. Then you have at least some statistics. Now you're like afraid to even check. You have a product that seems cool, it seems to work. Actually, I looked at your demo version, it's a cool product, it can do a lot, but you haven't sold it. All other entrepreneurs do the exact opposite. They make a fake product out of shit and sticks, then launch traffic there and see how much it costs to attract one client, whether this thing is needed at all or not. Understand? An alternative path of your development could be to make a landing page and an employee for sellers. Drive traffic there and see if anyone buys this dream. This dream, this is that you have an employee who gives you recommendations on what to do. Is there anyone who wants to buy this? So, let me try to offer a summary. Look, once again. From general to specific, I'm always trying to put it in my head and my company. I'll offer you this approach. First, clearly, in addition to a product spirit, there must be a commercial spirit in the company. These are two things that, well, like yin and yang, should complement each other in the most natural way, right? And what you yourself say, that you have to constantly clarify and ask again, trying to form an idea, I still doubt that I have a complete idea, so it's definitely a problem for the whole company. This means your people don't explain it better. And this automatically means that there might be problems with sales. This needs to be fixed. Second, I would actually descend to a level of greater generalization and once again figure out what exactly are the components in my asset set. That is, what is the nature of this expertise in artificial intelligence, in commercial commerce, marketplaces, how can this be mixed, and how a business model is built from this. Because, perhaps, especially at the beginning, you can look for other more promising options. Now you are directly facing what can be called neo-Luddites, right? That is, remember, this was in the 19th century, people who protested against factories. Now there will be many of them. And there are quite a few of them who protest against artificial intelligence. And, well, maybe you shouldn't hit this wall, maybe you can find a workaround. Since you are at the very beginning. Answer the question, what is the main component of my strengths and how exactly they are synthesized into an optimal cocktail that gives the greatest probability that the patient will want to accept it, and not reject it and try to fight it. What is happening to you now. So. And then, well, these things, these templates, figuratively speaking, test the market, really, trying to sell in different forms, piece by piece, as a set, to large companies, small companies, individuals, and so on, test the market. And this can be figuratively represented as, you know, like pliable dough, testing the market, it will take the desired shape if you are persistent enough in improving both the product and the commerce. Agreed. But I like a lot of things. Look, what I like about you. First, a cool idea, agree, right? Having artificial intelligence that tells you everything to do is cool. I would like to have such artificial intelligence. So, I'll ask for your help. I have a book selling on marketplaces. They all say, put artificial intelligence to improve it. Look, the product is cool. This is the future. In the product, we have two major uncertainties. First, how much of this functionality will marketplaces take upon themselves, and what will they leave to others. And the second uncertainty. We simply don't know what artificial intelligence will be able to do and what it won't. Every year, progress is such that you will have to constantly, constantly work on the product. And we have uncertainty here, right? You have a non-zero chance that everything you've done in your life will be flushed down the drain. But you're ready for that, right? But if AGI doesn't come in this form, then you can be a hero. I like that you have a basic business that supports your family. You just do consulting, services, and so on. I like that because if this AI thing doesn't work out, you won't be left with nothing. That's already good, right? You've built a base for yourself, now you have a chance for a unicorn. It's great that you have a base and a chance for a unicorn. Yes, of course, you've reduced the chances of a unicorn because you're also doing other things. Yes, you should only focus on this, there would be more chances. But on the other hand, the risk is very high, so it's normal. What do I like? The target audience, right? 500,000 sellers, most of whom are small with no budgets for a team. More than 80% do less than a million. When I think about business, I think, who would be bad to be one of these 500,000? You definitely shouldn't be like them. It's better to do something for them. You've already figured it out, right? That it's better to sell something to sellers than to be a seller? This is what I like, because you have many clients, you only need 5,000, and that will already be a good, solid business, right? This is what I like. That's it. Plus, your product works well. Actually, your product works better than other Excel spreadsheets that people use and so on. And in general, there's also a cool thing. We have the largest amount of data in the market, and we are forcing all our analytics service competitors, and we are stealing, well, we are taking their data for free, in fact. Yes, yes. There is no more data than we have. We are parsing all major sellers, all of them. And you say, like, what photos to take. For example, when my book just disappears from search results, my team finds out about it a month later. Yes, if there was artificial intelligence, it would immediately write: "Caution, something has happened there." In general, it works 24/7, constantly recommends. Cool, I like it. What I don't like about you at all is your ability to sell. This is a fundamental important thing, because, look, with the product you have, others would have already gone through three rounds at all seller conferences. Everything, everything, everything, you're not ashamed of your product anymore, right? No, it's not already, right? And you don't know how much one client will bring you. You just don't know. They seem to continue to use it, but your 17 clients, who were already your clients before, this is not indicative. You need to get from cold traffic, what is your customer acquisition cost, how much one client costs you. I suspect your customer lifetime value will be high. I think it will be 200-300,000 rubles. Yes, there, well, a couple of thousand dollars. Therefore, you can spend from $200 to $500 on one connected client. We are budgeting 100. Do you want to break even in the first month? It returns on the first payment. In short, what I did regarding traffic. You're right, we'll definitely do it. We, in short, have written to all marketplace bloggers on Telegram, YouTube, Instagram. There are bloggers here. We've written to all of them, asked how much advertising costs with them, integration. To buy out the entire blogger market, which responded to us, it will only cost about 30 million plus to buy out the entire information field related to marketplaces. So I believe I will have no problems with sales. Wait, from what you've told me, I see that this thing needs to be closed because it's not needed. Well, people, it's cool. There are many products that are cool. My social media gift service was awesome. Well, if no one needs it, you understand? No one needs it. Apparently, maybe people don't want to look analytically, they do it differently, or who knows. What you should have shown is that I talked to this many people, this is the most valuable thing you can do for your business right now, the most valuable thing is simply to show what you've drawn, the funnel, it's drawn the other way around. I talked to 1,000 people, yes, we have it like this, like this, like this. We already see it like this. And most likely, it won't be so beautiful. But in the future, we think that in the future it will be like this now. We are customer developing everyone we have in Sirena every week, and the product is constantly being improved. Until you gain experience and data from testing all this business and its feedback in the form of wanting or not wanting to buy and why yes or why no, then there's nothing more to talk about, and no professional investor will enter this. And strictly speaking, the prospects are quite difficult and vague, because, as Oscar said before, the recipe for success is always very non-linear. It is quite possible that you will build a unicorn. But it is no less likely that all of this will fall apart. For now, it seems that the chances are not small in this direction. The idea is good, the execution of the product is also not bad. But people just underestimate the role of sales. I am currently considering one investment. By the way, they even have to reach the stage of pitching. These are guys who are doing, as they say, Zoom for music education online. That's a very narrow specialization. And the first thing I said, after looking at their basic setup, I said, I'm ready to consider this if you immediately stop development and focus on marketing, because they have some good traction showing, but they are enthusiastically continuing to develop features. And I assume you have a similar story. This is a way to get distracted. This is a way to talk yourself and the team into doing important things and that something is about to start. But this "about to start" might never start. And, well, really, if you have some basic things that you believe can be sold, everything should be focused on getting feedback, getting sales or not getting sales, and drawing conclusions. And then we'll see. Well, that's it, it seems, and 100%, you can even leave one feature, only photos or only text, or, well, what specialization will be needed in the end, I think the more reliable it will be, because all universal specializations will be taken by corporations, for sure. Well, I think it will work. I think it will work. I like that you've already invested your own money, $100,000, right? The risk is that you will continue to develop your LLMs, go deep into the scientific process of how to do it, because here you can, well, you can endlessly develop, in general, buy traffic and see what happens. Yes, I think you need to find a landing page that sells your product. What specific pain? No, in general, this employee can do everything, this and that, and the third. Your situation is too complicated. Too complicated. Perhaps you need to go to a seller with one specific feature, and then everything else. Why is there rejection? The immune system kicks in because you come like this: "I'll do everything better than you." We had a guy at the last review who makes artificial intelligence for hotels, the booking department. He solves specific problems, they have surges, everything happens. And during these surges, they lose a lot of orders. And he solves a specific problem, a very specific pain, a large volume of requests in a short period of time, and that's it. And he's already selling 10 times more than you, having only one feature. And then, of course, they need to help these sellers. They, damn it, are losing all their money. We calculate the net profit of launching Sirena. It will show you in the red or in the black in 10 minutes. In general, Ranets, you're a hero. The only thing is, you don't know how to sell. Uh, there's something to learn. Either you find someone who knows how to do it as a partner, or you forbid yourself from adding new modules to your spaceship. And just, every seller should know about you. Every single one. It's not that difficult. There are only 500,000 of them. No, it's not difficult and not expensive. Yes. Yes, they gather at conferences and so on. Uh, show them the capabilities of your artificial intelligence, and that's it. Let's go. And next time you come to us, just send us your homework, how much it costs you to attract one client and how the funnel works, through which channels, how it works. Then there will be something to talk about. Okay, I'll write in a month. Alright, thank you. Thank you.