205. AI Startup Founded in 2024 Raises €15M Seed Round, Signs 7-Figure Deals With No Outbound Sales & Aims to Lead the Next AI Breakthrough Beyond LLMs Into Physics and Industrial Simulations w/ Miks Mikelsons & Dennis Just (Emmi AI)

We don't have any sales on organization, we are not doing our found at the moment. They slowly inbound into one of our researchers at conferences into our mailbox because they know that we're capable to solve big problems. And this is the best customer that we can at least at the moment, I would say, and that we can get, because if they know that they have a problem that they want to solve, they know that they have a budget, but they're very committed to solve it. I'd say that actually most of the knowledgeable disease have their own researchers in the team. Now already yes. But I think we've been surprised about the depth of the question and challenge, even at first. When the disease started, I think that there is shift happening now. And what happened with the language? If you can hold a large language model, now it's going to happen with these other spaces which are manufacturing engineering, science, which is much harder to crack, but obviously there is a shift. Even if you look at NVIDIA presentation, that's a keynote from Jensen. So he's also, this year, all his pieces about physics AI, and he's not so much about the languages and all that. You're doing entrepreneurship, I think what you learn over time is that there's, you don't, you never lose. Either you win in terms of experience, and money, or experience, and then you can utilize it in whatever next could you have. Hello, hello, hello. Dear listeners, welcome to another episode of the pursuit of scrappiness podcast. Whether you're building a business, running a team, or just starting out in your career, or vibe coding, we are here to bring you scrappy and actionable insights to help you become more productive. My name is Old Star Outcomes, and my co-host is Jan Zepz, I'm a friend. Productive as always, hey. Before we start, a quick reminder to follow us on Spotify and Apple Podcast. It helps more than you know, and in exchange for that you will find over 200 episodes already covering all topics. You need to become a scrappier and better version of yourself for your business, for your family, for yourself. So there is plenty to explore if this is your first episode. Plus by following us, you will be first to know when the new episode comes out every Tuesday morning. So open that Spotify, Apple Podcast app, click the follow button. And let's be friends. About today's topic, headlines are buzzing with the promise of AI for the past two years at least. Very intensively and for longer time already before. And they're buzzing with the promise of AI disrupting all kinds of businesses and industries. However, the most visible AI adoption for everyday users is, you know, people getting help with all sorts of individual tasks from GPs, asking some kind of maybe sensitive inappropriate questions and getting funny answers. And then actually solving real, real, real tasks, very helpful. But today we wanted to see what's happening a bit deeper in the revolution, more behind the scenes, not like something that consumers see every day. So we have invited a company that is disrupting a very particular vertical simulations for industrial engineering. So to uncover how companies are riding the AI adoption wave beyond chat GPT and the likes. So please welcome Mixed Meteor Sons and Dennis used founders of ME AI. Hey, guys. Thanks all those thanks, you know, great to be here. Good to see you guys. So yeah, in short about ME, it's an Austria headquarters deep tech startup developing next generation simulation technology for industrial engineering founded just in 2024 seems like yesterday. The company recently announced a 15 million euro seed funding round led by such known names as Speed Invest 3 VC Serena and push. And today we want to talk about building AI products selling complex products to large enterprise customers fundraising and some other topics here in dire. So let's kick things off with one of our favorite topics product development. So AI is poised to disrupt almost any industry and can do so in many different ways. And so we wanted to find out how you're thinking about narrowing down a specific specific specific problem to solve, you know, to ensure. Early product development focus and probably also going forward because you might think that oh my god, we can do everything with this with this beast. So so how to how to really kind of narrow it down. Good, not sure if we have the perfect answer that everybody can relate to right I think maybe tying it a little bit to the founding story of ME and. Where we are in this journey because obviously the journey to build building great product and stuff so we've. We've actually got thrown the problem at us I would say and we are currently trying to build a solution around it. So what happened maybe to go 6 to 12 months back is that our third co-founder Johannes was one of the hats behind Aurora, which is one of the leading mid sized weather models at the moment and the first true. If you want to call it foundation model for. Yeah, physics or astrophysics and you basically took that knowledge went from Microsoft back to lens. That's the reason why our companies had ported there and had the intent to took what take all the learnings from the time at Microsoft and really solve the problem in industrial engineering around utilizing AI to essentially solve or accelerate big problems. And this is really what we build on at the moment so I would even say it makes correct me that we don't yet have a true product. So we very much have an architecture that helps us and big industrial companies to basically use AI to simulate true large scale problems from I don't know airplane aerodynamics to power transformer. Very more electrics to molding behaviors of an up you from for example so it's very much physics phenomena agnotic and the point where we are is we do have that architecture we are currently in the first step of the productizing it and serving first customers but we don't have a packaging we don't have a product that people can test right so it's very much. It's part of the scriptiness that you have to do right as a startup but this is where we are and we're trying to figure out with all the knowledge I think that all of us gained over the last 15 years doing business. So you're basically still in the process of narrowing it down as well somewhat somewhat yes I mean that there are really strong hypothesis that you have right and I think we are very much a I would say half research have product company. On the research side we published quite a lot of papers and I would say that from a tech perspective we do lead in the physics I simulation field from a deep tech perspective at the moment. But what would you have to I think record nice or we learned as well over the last six or 12 months is that we cannot throw our Python code to Airbus and tell them gosh you know guys train your new aerodynamics model based on our architecture because they don't have the knowledge of how to deal with it where they don't have an in-depth understanding of how to utilize deep learning in fairly I would say old school numerical driven processes. And this is all what feeds into this process of okay we can technically at the moment server customer by helping him to build these models and then implemented in their workflows but we're not yet at the point where we can give them a package that we can just put in you know their software workflows whatsoever and it runs. But this is the very much the step that we obviously have to make and it's the bet that the investors are putting on us as well with the 15 million which is not a small amount. And I think maybe just to add also what you can hear from Dennis here is that we are aiming here to work really with it but I would call high impact customers so that's really our customers that are truly large in types of the scale and they have big problems that they want to tackle with the AI for a lot of them at the moment it's a bit of like a codec moment meaning that they know that they need to reinvent themselves if they don't do that then somebody else from their peers will do that. And so they definitely they need to put the part of this game and then what I think where we become interesting for them is that we. I think in the A& research show particularly there is a lot about that street credit that you have and if you have a ability to really have a true real groundbreaking research that obviously is not then coming with a turk of how the Johannes and the team around him then then we really are able to pick up these opportunities also when yeah connect with these customers like Dennis mentioned. So that's also how we are able to get closer to these seven digit ACVs because essentially we're working with really through global players that has a challenge with their hand they need know that they need to reinvent themselves but they don't know yet maybe where to start and how to what needs to be done first. And it's not that we are doing only the managed service here but but it's really that maybe the starting point for where we also can then benefit both us we can get better know industry understanding the specific domains be the thermo electric be that aerodynamics cars planes you name and then from there to pick up also where to do you go deeper maybe. So that would be maybe with broader answer from from your question and maybe to add there and give you a bit of a for everyone as well as listening a bit of a broader picture I think what we see I mean AI is bas in high might some more question but AI is around for 20 years already it was just called differently before. So there's technically nothing new I think the state that we're reaching now what did we let's say that we reached to three years ago with GP to two and that we are trying to achieve in physics now is that you come at a point with AI deep learning how we want to name it that you can really solve real world problems with it. And in our time to our case or simulations of physics AI they have been PDE so kind of small physics models as well around for like 20 years but they have been very much focused on small scale problems so I don't know. Simulate me away and try to get you know the the turbulent behavior of a wave versus the real need to make an example to get an understanding of how ship behaves in the waters you need to simulate the ocean you need to simulate the metal as well that is in the ocean with the ship and how it kind of moves and if you consider this is the problem you're at the point where I think till now it was not possible to simulate that with the AI. We're not even within a mile because I mean even on super computers it's not working as I don't know 300 million match points multi physics phenomena and usually you simplify the problem extremely to get at the some sort of an of an understanding but now with the capability of new operators and parallel processing etc. You at all haven't the ability to simulate this as a whole and I think this is what excites a lot of industry players at the moment because they start to acknowledge and see that we can solve problems that they haven't been able to solve before and with that either unlock new business that they haven't been able to touch before or unlock an acceleration of production processes that haven't been changed for the last 20 years. And this is essentially the opportunity that has anywhere to think. When you mention simulating the ocean and similar and just immediately start a thing of video game development and how can that because because you know there's similar games that seem to have like a vast vast world in which it operates but it's still it's finite right. It will somehow very very maybe with your technology we can have GTA 6 finally exactly where truly I think fun fact there our CTO is was form a city or black shark would get the simulations for Microsoft flight. So the technology I would generally say is not too far away obviously they are serving different cases. But I think it's also important to highlight and as I listen here to Dennis always remember that I think. So we were on the fundraise journey to the Silicon Valley as well and as part of the trip we visited the answers of this answers is actually a company that is. One of us you call like incumbents right so you maybe never heard of them but they are used by our majority of these big large enterprise engineering companies on even smaller ones to really run the simulations which what's called this numeric simulation so numerics is. In a very simple five terms you can take the classic you know there's I think you call them in English integral formal is basically that you will learn at the university and just run them and try to solve different algorithms basically to to and get sold your new and a car design from the all the physical problem perspective. An answer is providing a lot of yeah but both they have a platform they also provide what's called is ground through it against you what can compare the simulations. But when you go to them in person you see that it's clearly it's an old industry they have this their offices like from the serial office from the 90s you know those that you can see the everything in grade box tables corner officers are for the top management. And nobody is there actually because nobody wants to see that everyone is working from home and but the same time company is valued I think this was at 35 billion acquisition last year that happens at 35 billion views dollars for a company. I should not just acquire them and I still think they will be the ability for the final they got the regulator of approvals but I think it's pretty much done deal by by end of this summer. And I think that just the describes that there are like few few increments which are actually not able to serve the changing shifts that you can see now. And definitely they also would be interested to tap into the AI potential be the same as our customers here they didn't need to reinvent themselves a question is whether they can be fast enough. Well given your given your answers I also need to be to reinvest our conversation questions and direction so. So let me clarify this because you know you mentioned these you know six seven figure contracts that you have secured at the same time you are still productizing. So are these like kind of discovery contracts that we're going to be doing you know R&D together with you to to kind of find solutions to X and Y problems or what kind of like what kind of deals are these. It's part part so with so we have two big customers at the moment they're both seven digit contracts and we had to prove ourselves before us usually in the I would say enterprise field in AI when you do deep tech solutions you go through a pilot phase which is true to three months while we get a little less money you know after the pilot you kind of step into a year to be a contract and whatever it is and then what we do is mix professional service where we usually build a joint team. With the industry partner and then yeah define the scope usually it's tied to with some sort of foundation model development that they can implement in their workload. I think what you have to acknowledge trying to the question is that some of the things that we so it's not that we have a solution we're just implementing it for them but there's still a lot of research happening. Because the problems that they're trying to solve are so big that nobody has solved them before and I think so this is the exploratory research part where the contractors will have a license fee for obviously the technology and for the maintenance of the models afterwards. So I would say and therefore this is this mix of research and product at the moment so very much in the middle even if you look at the contracts where part is doing joint research with the companies where we get the simulation engineer. Let's say industry know how from the partner that we're working with and usually the learning know how comes to mass trying to build a foundation model together to solve the case and then implemented into whatever either software or process that they have. So this is very much where we are. How does one find these these customers I mean you guys not exactly from this industry before right so so are there so few of them that it's easy. Now the cool thing I think in our case is we are very much as we're coming from research and we are publishing results and they find us. So we don't have any sales on organization we are not doing outbound at the moment they solely inbound into one of our researchers at conferences into our mailbox because they know that we're capable to solve big problems. And this is the best customer that we can at least at the moment I would say and that we can get because if they know that they have a problem that they want to solve they know that they have a budget but they're very committed to solve it. And the other type of customer that we as well as see which is still interesting but not short term is this the company 10 billion I want to do AI executive XYZ told me I have to do AI tell me how you can help us. So these are very much the two type of organizations we're seeing in a sales process at the moment whereas the first one obviously is the one that you want to talk to because I mean we're 20 people organization we can only serve in the current setup likely three to four customers because it's quite an effort to engage with the customers and we're scaling the team want to have more customers in the future. But yeah it's it's very in the own focus. Yeah actually what you described is this perfect sales sort of match where the company knows they have a problem they know they don't need to be convinced and they need a solution right. And so often sales is like vice versa where you have like I have this cool product which will make things faster for you and more efficient for you. But you talk to a company who is like you know we were other priorities at the moment you know I trust your tool is good but like you know what I but with the scene bound like I'm just thinking what lessons can other companies in the field you you need to. Yeah you need to put out research you need to put out publications you need to produce content how is there something they can learn from you. I mean that this is our way which we because we are still very research having what we're doing today and we release foundation models ourselves in and foundation architecture is just where we play. 15 or 20 employees are researchers right so this is where they feel comfortable so we feel comfortable comfortable as a company to basically use this channel of publications. And I think what I'm not coming from the space I think mix mix is well not I underestimated this before so actually there are a lot of people reading papers and there are a lot of people as well or researchers in big organizations looking at what is happening in research and then trying to utilize that to solve the problem so I mean at least it's one of my learnings is that I historically underestimated the impact that you can make with papers or with research conferences. Because obviously it's very tech heavy but once you have a solution that fits into this type of profile that actually works very well because you can build a lot of trust. So if they believe that you are the right company and we have one case with our first customer so they really chased us to become one of their first customers and that's I mean that's what you want to have in the end and feels very good if you can solve the problem. I think when conferences is important also maybe it's very funny for the listeners to understand that the conference here is not a classic that you go to Germany to Munich to the big fair and then you have a lot of sales exchanges there is actually the conference where you go there is it's 80-90% researchers of big deep minds search researchers that come there with a poster like literally printed out big poster that they put on a wall and then they gather people around them and just explain. How did it tackle a specific problem and what solution was created and to that when you actually can generate and pick some quite good leads also in my conference where you don't come out with your classic you know hey here is my link in account or business card let's talk you for a coffee. Of course part of that is happening but I think this is so much differently organized and my personal discovery also that the science here really is taking really far in terms of getting customers soon as well. But I think it also makes a lot of sense and also for other companies in deep tech meaning that especially if you're dealing with customers who also have big engineering scientific teams right you need to convince them that you can do stuff that they can't and that they you know you are smart and you know what you're talking about and you know a sales meeting might not. Do that in a probably also a nicely refined product also might not be enough to do to do to give that conviction so also. But is this one thing you have done the right direction there maybe to give one caveat to the that we always have internally discussed I mean as great as this from a channel perspective you have to give something out to the people from your core IP. Because they need to understand the solution that you did at least from a structural perspective right doesn't mean that you need to give out the source code of the solution or the architecture that we have. But if you're not willing to share and discuss part of your resolution for a problem with the community this path doesn't make sense and this is always a bit of a debate as well even with investors internally that we have what where the question is you know. Does it make sense to use this track because you you know you will give give some input to your competitors that they might not have over the next three years to accelerate them while obviously you stay ahead and build this perception of you know thought leadership. Which is a very fine line to it even in the publications that we're doing it always the case you know what do you publish just a paper do you publish the model you publish the forward part of the model that you publish the source code to what extent. But so I think with this as a general that there are a lot of fundamental questions that from case to case you have to discuss and then evaluate for yourself. Well I guess if you're able to create a product that is black box and speeds out very convincing results then you don't need to share anything and you don't need to get on these exploratory journey commercial relationships right but it. Then then you probably harder to secure such contracts having been founded in the less than a year ago or a year ago or so so. I think I think you can look at us I mean just to illustrate I mean a lot of people are when hearing or story there are you're doing volunteer approach and in a way that is last case that we want to achieve here is that we actually go through this high touch motion at the beginning and then we had it over back to the customer to use and then. Of course we are available or if necessary going forward but but then it's more about reoccurring charging maybe in licensed model shifting to that towards it so that's that's that's our aim and goal and we're also I think pretty clear with our customers that that's that's how we envision going forward in this. And being honest and I think if you look at the summer the big players which are already becoming a household names in the space not particularly doing physics say I but more like I don't know name it Mr. A.I. If you look at would look under a hood they actually are doing quite a lot the same because it's it's difficult to really generate a real revenue at the moment in the eye space if I don't if you are not like you know. And of course that GBT cloud or whatnot I mean still the investment that you need to do in terms of both getting the right talent and getting the market to his ability all the brand awareness what you can say and it's really it's really that you need to have also of course these big customers be contracts in and they will acquire quite quite a big of hand holding at the beginning. What about what about investors how do they how do they see this this model how do they how to convince them that this makes sense that you know historically this kind of manage services doesn't count as revenue and our professional services and and this kind of does the A.I. I would generally say that investors get more and more comfortable to to have some sort of managed service professional service within your industry and I think specifically the Palantir example but there are other ones like Mr. for example from what I know they make most of the revenue with manage the rest of the licensing on that as well for enterprise at least I think makes them comfortable. So obviously the I take is one and secondly if you look at the industry at the moment so specifically the problem we're trying to solve with around physics the eyes there only I would say two to three companies in the world who really have their own core technology meaning core architecture which they can train. Models on and solve we were problems with everybody else is just going at you know the standard architectures on top of Nvidia GPUs which is physics name or Domino and then basically utilizing that that has a lot of limitations. So in for investors in our case I think one they feel comfortable with the approach because once you have a customer it's actually a lot of money and a lot of cash flow and you're getting very much locked into the process. So there's it's it's very hard and to exchange the model in a workflow I mean as hard as it is to to change an America simulation process and into I think the ability to really create your core technology that has used P or differentiated even against in videos of the world. Okay but like does it like is it easy to easy or hard I mean it sounds like very complicated stuff so is it something that also takes a lot of effort to actually explain what you guys are doing or are you just showing that there is this big pull effects from from these industrial players. It's a real problem that needs to be solved and then you know VCs don't really have to have actually physics to understand. This is a to add to this question how does it go you go into the room with this VCs how how in depth they they they ask you probe you about questions or or are they in between their golf trips and and drinking apparel spritz to just writing checks to AI companies like how deep did you go in these discussions. So I would I would say that actually most of the knowledgeable VCs have their own researchers in the team at the moment. So the question I mean they're different types obviously they have their clearly deep tech VCs we have a bit of a deeper understanding of the field and they are they are more generalistic you want to say this way but I think we've been surprised about the depth of question and challenge even in first or second calls. I think what I wouldn't say made it easy for us to raise but a bit simpler is that one there is a big hypothesis of species at the moment on the few because it's an untapped market and it's an opportunity that is unquestion right so if you can solve AI for engineering you will have a market you have definitely a trillion dollar opportunity in front of you and you can go at it. So this helps and this even creates VC and down towards us even before we raised around and announced around and I think second to your how deep they go day I mean at least in our case we've been quite grilled on the technology. I think the good thing is that with the publications and the references to the publications we've done again time to what we discussed at the beginning right you have a lot of thought leadership and you have a lot of trust to them that they actually believe you. With the argumentation and the products that you put out even if there's not yet a true product that they can download and test. Yeah and maybe to add to this one I think my version to this one is that a they're definitely on the VC side they recognize that there is shift happening now and what happened with the language if you can call like a large language models now is going to happen with these other spaces which is manufacturing engineering science which is much harder to crack but obviously there is a shift. Even if we look at Nvidia presentation that Akino is from Jensen Densen so he's also this year all these pieces about physics AI and he's not so much about the languages and all that. Second I think for us or my personal takeaway is that if the VC is not able to get from the first touch point to the second step already where they bring in someone who can really ask the right questions then naturally that's an in signal also for me that that's not the right investor in okay. So yes there are cases where you clearly understand that is hard to explain the story but I would say who does to us. I think we have been able to give craft the story so that a person who is also further away from the technical details can understand it. But then when you don't get to the second call which is going going more into depth about the research side the technology actually does a sign also for us that this is not the right partner to be having a board meeting later on. Were you in position to have to filter and say no to some investors or was that not up to you. Like luckily yes so we've been I mean I think the hype terms over subscribe right or over interested. We've been at this point I think we try to really pick the investors that understand the point where we are. So there are a lot of we see that you don't go at this are now you have product market fit let's go to market go back. I think that's usually how I probably I would say usual hypothesis or we see I think for us it was important to have a bit more time. Because we know that this GPK 2 movement is happening in next 12 months and I heard that we're in and we really want to crack it. Because we know that if you are the first ones to deliver it say foundation models or reference models for the case that you're in. You can utilize the momentum going forward and this is really what we want to kind of accelerate. So therefore we actually don't go all in on earning money at the moment. Although we I think do reasonable revenue this year for a first year startup. But really focus on cracking the last problems of our field to then productize afterwards. I hope you have some good template emails to say no thank you. It's not that we have to say thank you for it on a 10 investors I think but to us we we definitely had to say no. It always feels bad right I can tell you from a process perspective. Well at the same time it's the right thing to do and to just focus on the people that you feel comfortable with. It's one of the first times actually we found the proof that it's actually happening. VCs we talked to VCs as well and they always say we have to compete for founders we have to always compete. At the same time we talked to founder and they say like fundraising is very hard and I always thought like who's lying. No I think it's tricky. Yeah and I think after beer in mind obviously I think there is this space as AI of course is pretty high at the moment and you can question a lot like when this hype will go down a bit maybe. But I still feel that yeah there is the right opportunities find the right also investors at the moment and that's what you can clearly see. But that's not getting the first stretch at the moment maybe the recent weeks that you can see some fundraising announcements also here in the Baltics. You can question of course but I think that I mean to me personally it's being a good story for us that we have been able to really find the right investors with the right story and also able to then partner for the next stage that we are trying to unlock here. Well this error seems like last two years with AI of course over hyped as well sometime but it does seem to you know you guys solve some industrial problem. We all know it's a big big market what we had a few years ago I guess in globally maybe not so much in you know our region but there was like you know we're building images of monkeys and selling them and you know or something like virtual world which probably has potential but like there was there was some interesting funding rounds in these spaces for sure. Great comparison I mean what you see at the moment is that you're somewhat back to and I don't know if we feel it a little bit and get from the investors that you're back to the under XAR at the moment while at the same time to to cover this up from an argumentation perspective and I think this is what you kind of into two years as well as we're just at the starting point of creating impact into the world and we already see that you can monetize it right so. Obviously the question is can you build a one billion dollar revenue company out of any I think potential is there will you get there in the next 5 to 10 years like we don't know either right we're trying and but we don't know either. But I think this is a bit the reason why everybody's jumping at these verticals at the moment in AI because they see the chance is now to unlock these verticals and there will be a winner in the verticals which will not be. Open AI or master because they are very much tied to language with you and vision and for them it's super hard to go left at right and I think this is the reason why you see all these funding rounds in biology right in. Agentech in physics in chemical in a chemical field or protein field because it's it really unlocks potential that hasn't been done before all of them have this monetization challenge right that we face as well right first proof is there is it scalable and repeatable question mark but if you hit this point then you go to the hundred 200 XAR radiation that the moment is kind of crazy if you reflect but I think. It's great to see at least that there's interest in the space and that there's impact that we can create but I think that's how hype cycles work and how also bubbles work that you in order to find the winners you need to also get burned on a lot of a lot of well I wouldn't call them losers but let's say them missed missed opportunities so. I think it's just the natural way of the world and I'm not sure that you can really do it in a better way is just you just need to you know limit the grief and protect the people that need to be protected but otherwise you know if it's if it's professionals doing doing these bets then you know that's their job they should be doing it and and some of them getting it burned and that's the way of the word. I agree but at the same time you describe it as an investor dilemma I think it's a bit of a founder dilemma as well because I mean I've had this in the past what if you just become the number two or three are you happy with the outcome as a founder or do you feel as you call it loser right mentally I think you know ultimately ultimately you need a company with the hypothesis that you have around product market you know cultural market people in the company. It doesn't mean that you win and there's a lot of luck involved in the process as long as you lift up to let's say the best hypothesis in the best effort it the outcome will be fine yeah I think you know there are of course these exceptional perfectionist you know like Olympic gold medalist professional athletes is like you know number two is is never good enough and also in entrepreneurship. But I think that you know most people will find meaningful work and and building something valuable and you know still creating generational wealth. It's a very fulfilling journey while very tiring journey etc but like I mean if you don't go bust the second year after or the first year after raising then you know even if it doesn't work out even if you not only don't don't become a number two or three company but even if you do not make it after you know companies fail after 10 years after 15 years. I think there's still value to be to be found also in these situations for the founders and do take some money of the table on the second there is as early as possible I think that's something that we have learned from from a lot of these conversations that. You know it's not like you're gonna stop working but but if you're if you're not gonna do this and you know five years down the line 10 years down the line you still haven't taken any money of the table and everything goes goes bananas then or goes belly up then then basically it does leave you with a lot of positive experience but but not much financials to show for so yeah different different outcomes for different. Now we've become a startup or entrepreneur advisory podcast but I 100% agree I think this is what everybody should do if you have the opportunity right and I think there's you're doing entrepreneurship I think what you learn over time is that there's you don't you never lose when either you win in terms of experience and money or experience and then you can utilize it in whatever next kick you have. And you know monetize it or try to monetize it. Would be cool to test VCs reaction you know you're you're doing this sought after projects you know you're getting inbound you're at seed stage have like one or two contracts and then like hey can we do a millie here on the secondary like we've been doing this for nine months already you know. I just see you know just to see like how how crazy the hype is it all like but I think you always have a lot of new investors to to to force it a little bit I think doing this with existing investors will be tough they will very likely just smile at you and then hang up the phone quite likely quite likely. One last topic we wanted to cover we spoke about it a lot when we started in covid times and after covid and then it kind of became secondary nature but now I think with with a lot of companies taking a loud stand especially like big tech you know come back to the office like five times a week or seven times a week or. I heard this yeah exactly nine and six that was the one right so from nine to nine six days a week is that the translation of nine and six so yeah so so we wanted to talk about like how you're organizing things because you know it starts with the founders you are already right in from different countries based in different countries so kind of what is your philosophy how do you how do you go about. You know organizing this yeah I think I can maybe take first this one from the practical perspective yes you write I mean we are distributed out of the by disability not DNA that's how we started from our perspective with agreement that we definitely value these people being together in moments so we need to show up at least once a month together in lens that is our main main habit the moment in Austria. And we gathered everyone for five days once a month the whole team needs to be there and really then work under strategic moment planning sessions maybe get some retrospectives have really alignment between the teams that we have inside and and then we can go back to our tables and desks and continue to work remotely on a day-to-day life. So kind of sounds easy but in practice of course it is a mindset that you need to have to have to embed in the daily life you need to also make sure that people that are maybe previously that haven't been working in the most the thing really quickly on board so they set up and are more communicating in day-to-day life than they used to be doing this by sitting alone and working on the code or and doing their research in a closed box. I think for me and then is with the background of localite which is fully remote as well and it was nothing new and I think we just took the best playbook from the previous and started to play here which is a bit more I would say like a hybrid approach right so we are not it's not that we are not saying no for in person moments that because in localized case we maybe had the opportunity to meet in practice with the entire team maybe once a year maybe twice a year. Here we are really practicing these 10-12 moments a year together and then we can go back to the whole desk. Second of course I think is the reason that we recognize obviously about yeah we want sounds really very classic we want to guess the top talent but I mean for us as a new player in the block we need to also have we need to have this flexibility right so we need to get people from different parts of Europe and make sure that we are able maybe then get to the talent that doesn't want now to go to Paris right immediately and to be part of what not Mr Alore or meta AI teams there or the same in any other big center but you can see in the aims of AI in Europe so we are also a good opportunity for everyone to join our team and still stay close to where you are in the way today. Would that change in the future into some kind of different hard approach or something I mean the future would show but I think for us it's really working well at the moment and allows us to really be well connected well being distributed. And maybe to put it one I think argument is a counter poll towards is not nine six as well out and I think tying to the hybrid approach I mean ultimately we are definitely work horses but we're human at the same time and we have family rights I think all of us have our wife kids grandparents we do live in different locations or different countries even so you need to find some sort of a balance between the both. And I think this is very much the outcome which is can never be black and white but where we felt like with the localized experience and and the fully remote approach and all the goods and beds from that we wanted to shift shifted a little bit more towards on site I think ultimately it's clear that being in the same office for all of the time as a full team is the most efficient way how to work. But then the question is do you get the right talents in the place where you are and can you get all the people into then for us the answer to the last one is no. So we needed to find a solution that kind of works in between for everyone where we felt like we need to see person moments and we really want to to mingle to have copy breaks with the team so let's force them to be at the office one week a month. Well at the same time for three weeks a month everybody can work in the same time zone for wherever they want and this way we have the ability to hire basically great talent in all of the views that we are having at the moment. Actually it's what a lot of companies try to get people to gather those who are remote partly or fully just practically from a perspective if you compare costs of just let's say I'm having a full office everybody in office in an ideal world if they would live in one city versus flying in for one week. So everybody is there is is it more expensive to run like this less expensive or maybe you know I think it's more expensive because you need both setups right so you need to office because you need you meet in person for a week class you need to fly people in I think practically what it means for us is that it's roughly one came more per person for month in terms of travel costs that we encourage but then I think looking at it from the other side. Is it worth 1000 euro per month invested money that people meet themselves five days a week I think we can clearly say yes to that at the moment and you don't have hundreds of people as well so not exactly yeah yeah. And then if you maintain this research or ratio if you if you scale like five 10 x and still have the same research or ratio. And it this would not be the same ratio going forward but I would say until the end of the year will keep the ratio and then it will flatten because you need more software engineering at this point and more product and more design which at the moment right if you look at our website or the stuff that we release I'm always coming from small PDF and local lives where we had very big. You know user experience research product design heaviness sometimes it hurts. At the same time yeah it's it's not what we stand for it yeah I think maybe adding just your question also what always have in mind is that everything that we try to build here is kind of future proof I mean from day zero day one. So I think it's it would be naive to think that we are able to open now an office even I don't know let's say Berlin and then to have there for next five years everyone in the Berlin only hiding we anyways would have encountering critically this problem of hey we have the great guys now in Paris. There is where people are not helping you eager what not and what what what would you do then the open five hubs and you anyways need to maintain that they are like somehow connected together that they have these in person moments. So I think we are trying here basically from day zero building for the future proof in terms of being distributed. And what you what you said is actually even like that at least in your case everybody is the same right you know you when you have one big office and then you would hire one guy in Helsinki one guy in Paris these people would feel like you know am I given the same opportunities as the rest of the team that hang out and drink beers every every Friday right so it's that that's even harder maybe. Exactly so I think it's something I think what you don't want to repeat what basically is getting into the craziness of having you know 50 different locations different continents different cultures completely I mean I think we have learned some some some lessons from localized times on this one so. Everything time being quite cautious about yeah and not to you're having something I was a we are vulnerable on European really for the moment that's what makes us maybe different from others as well. All right guys great chat very exciting stuff we're going to be following your progress so do report do find time to also take care of the PR side not just deep research. And it's the same. Yeah kind of kind of very particular type of PR. You should have a look at super intelligence website the guy from from open air who raise I think what is one billion and still having one year after the page is just. No website no website basically and yes so I think this space works differently these days busy researching all right guys so so thanks a lot and to the listeners until next time. Thank you was a pleasure thanks guys. If you like this show remember to leave us a rating or review it helps other people to discover the pursuit of scrappiness.