AI Billion-Dollar Founder: AI Will Replace Most Jobs – Nikola Mrkšić

0:28Is the company profitable? round they've raised 50 million dollars and are valued at over 500 million dollars in 2025. polyai also ranked eighth on the sunday times 100 fastest growing tech companies of 2025. you unlocked institutional money just by presenting a great idea by three cambridge graduates honestly i don't even know if it was a great idea right how are these valuations done scarcity right scarcity of talent where you know that talent not idea talent talent ideas were nothing. And overnight, we clicked a button and we started taking 50 ,000 calls a day with our system. Now, the average contact center agent does about 50 calls a day.

1:08So this system click, it's there. It does the work of a thousand people. Now, that is why you get the upfront investment and the bets. If you can't communicate to the world why your software is going to be huge for them, then your software's probably not great for anyone. There was just a year where we were like, we're going to take all the sales calls between the two of us. I've never done more work faster and better than that. What you need is a problem solver who can continuously feed the information back to the product and evolve the company. Do you think entrepreneurs are made or born?

1:4434 years old, worth$50 million. Very few 34-year-old people can say that. How does that make you feel?

1:59Nicola, you co-founded Poly AI, which is now worth over 500 million dollars. What's the company's mission? To change the way that people feel about voice assistance, to bring about a future where we speak to machines as if we're speaking to each other, and where the world is, a different place where you pick up the phone, hear a different ringtone, you go like, oh, they've got AI that works. Amazing, let me just go through it, rather than cursing the day you were born. So I hope that we get to the point where our technology powers over half of those interactions you're going to have. And you're going to be having a lot more of them.

2:33Can you give me one particular use case that your service is addressing? We work with OpenTable with a number of different restaurants and pubs in the UK to take reservations, right? So that's a simple, easy to understand use case. Everyone's done it. You call, you know, people in restaurants are busy. They may pick up, sometimes they don't. This thing picks up immediately, you know. takes your reservation, you move on with your day. That's a simple use case. But we go into much more complex things, you know, like technical troubleshooting, order tracking, returns for retailers. Really, whatever happens at large scale in contact centers, we automate.

3:17Fascinating. I want to go back to your childhood, particularly your late teens. how would you describe yourself because you are a tech entrepreneur and i think based on hollywood a lot of people have a very particular idea of how a tech entrepreneur looks like so how would you describe yourself yeah i don't know if i fit that stereotype in some ways i do so i'm serbian went to a highly specialized math school not so much because i went to a highly specialist math school yeah yeah yeah we're like from what age um so that would have been from the age of 12 13 kind of like in serbia that's late late elementary school and then all of high school yeah and is that something that your parents chose or you just want oh i insisted you insisted i insisted i competed in maths as a kid and you know i was a problematic kid i in that like i would just talk to anyone in class i would disrupt every class that i was in i was a nuisance i think in spain we have different definitions of a problematic kid.

4:18I think at the age of 12, being good at math and wanting to just... No, no, I was disruptive in class. I was kicked out of every class just because I'm an extrovert. I talk to people. They put me next to the biggest introvert in class and I would turn them into a chatterbox, right? But you were still good at math. So what they started doing was they would have this small room next to the room and it'd be like, go there and I don't know, do math. So I spent half of all the classes just solving random math problems. And then in the first years of those competitions, I was like number one in the country.

4:49Later that changed and I was nowhere near as good. But it was just, I had a knack for it. And you know, like talent is like where you meet like, you know, you're good at something, then you start doing it more. Then because you're good at it, you do more of it, right? So it's always hard work and talent. And it was just like where I think I had the highest ROI and effort, right? And Serbia was an interesting place when I was growing up, right? We had the breakup of Yugoslavia, wars, sanctions, hyperinflation. It was a tough time. My parents' generation saw a decade where they lost everything. You had a million refugees in Serbia from Bosnia and Croatia.

5:28How did this affect your perception of life? I think it became very important to take every opportunity in front of you and realize it with extreme determination, almost desperation. so for me and a lot of my friends that ended up going to that school that was a way to just get to a better life because if you go to maths and engineering and this is a classic immigrant thing in a lot of places where if you go to maths you're going to get scholarships you're going to get into good universities and you're going to get a job that is secure pays well in many ways I've come to realize that what people see in a job at Google or Facebook today is very similar to how my parent generation saw a safe job in a large state enterprise.

6:14It's just safe, pays well, gets you a good life, enough money for you, your family, maybe summer holiday. Is that what you were striving for when you were in your early teens? I think that's where a lot of our parents guided us to and the schools and everyone, right? Where it was just a matter of be good at maths and that's pretty transferable, right? And then, you know, you kind of go into like your classical engineering disciplines and you're kind of sorted, right? So it wasn't like you've always known that you wanted to be an entrepreneur. No, not at all. Not at all. I think if you asked me, like for the longest of times as a kid, I wanted to be a neurosurgeon.

6:50Never really pursued it, right? And then, you know, like if you asked me like what else I would have loved to be in life, it's probably a diplomat. I love politics, I love history. But this was just like, like your interests are great. What you did, if you were at all good at maths and stuff, is you doubled down on that because that's a way to get into strong foreign schools, to get scholarships, et cetera. It was just like... So that math school is what got you into Cambridge? Yeah. Walk me through your experience at Cambridge. Well, I mean, honestly, I don't feel like it was a change for my high school at all.

7:24It was just an academic pressure cooker, right? I think that American liberal arts... Did you find it easy? Not at all. It is hard by design. It is very competitive. And you're a subhuman if you're not in the top quartile, right? If you don't get a first, you don't get two, you know? Did you get a first? Only in my third year, which is the one that counts. But starting off, I was arrogant. I thought I was very good. Were you very good? I was good. I think everyone there is very good. What defines good at Cambridge? So it's just a ranking. Like you could bring... Good grades. Well, let me tell you this.

7:58First year class, the guy goes, how many of you of the 200 people went to the International Maths Olympiad no I didn't a lot of my friends that's the first question he just asked that right out of interest yeah half of the room raised their hands so at that point that was the collection of which degree was this? this was maths maths degree yeah and you know you go and you realize damn this is gonna get feisty because it's not easy to be in the top quartile if you're not that good so that excites you it doesn't oh no I hated it I thought it was just like I was like how long Why are we doing this?

8:31Like, how do we define success? And it was unclear that that was done for any reason other than because they could and because they thought they should, right? And I think a lot of the British education stems back from, like, you know, there's a famous expression about a battle. I don't know if it was Trafalgar. It was won on the fields of Eton, right? And, like, how they compete in sports and stuff was the, like, imperial ethos of, like, just fight, do better. you see it much more in america today with how they like go and do sports and every kid in america does like you know american football and has had five concussions and i'm like but you weren't even like close to like getting professional and yet it's like a point of pride like hey i had like five concussions yeah and it's like i always wonder why and i think in many ways you know my eastern european nerdy version that then flowed into cambridge those years for a lot of europe was just like you compete in maths.

9:23You treat it as if your life depends on it. And I don't know. I don't think. I didn't enjoy it. By the end of it, I wanted to go into finance. I was rebelling against the whole thing. Almost ended up not doing what I do just to do something else. Does going to Cambridge help you become an entrepreneur in any way? By presenting opportunities, for sure. At the end of my undergrad, I was about to bail on the whole thing and do investment banking. so my master supervisor goes if you don't want to do a PhD at least speak to this guy who just left his professorship to start this company and I go and meet this guy and his co-founder another Cambridge professor goes like don't be a PowerPoint monkey here why don't you join this company but I'm joining this blah blah blah he's like what if you did a PhD with me while working for the company and I'm like damn that's like an offer that's unique you get it because you're at the right place at the right time and you did everything you had to be good enough for them to launch your day.

10:20That was a startup. Yeah, and I was like, okay, handshake. And that's how I joined that startup. That startup ended up acquired by Apple. What kind of startup? So we also did voice assistants, right? That was like that whole lineage of building this technology comes from the Cambridge Research Group where I did my PhD. But your background was in mathematics, not computer science. Computer science and maths. Computer science and maths, okay. Yeah, yeah. Same thing. Especially at Cambridge, this stuff is so theoretical that you barely feel like you're even doing computer science. You're just doing a different kind of maths.

10:53So, okay, I can see a clear common denominator in where you got inspiration from. So you're very good at math. You get into Cambridge. You get also exposed to computer science and math. At some point, you have doubts if you want to continue in that career path, but you have this one person that gives you this one-in-a-lifetime opportunity, and then is when you get exposed to entrepreneurship. so you see that a company can be built and also sold. Yeah, yeah, yeah, yeah. But the building part is what excited me, right? The building part. Like, you start working for a startup. Yes. As an early hire, you're basically a co-founder, right?

11:31If you want to be. Like, you can, maybe the perk of being like that kind of semi-co-founder is that you can drive as hard as you want. Yeah. But you also maybe don't have to at all times. That I learned later when I started my own thing. But the ownership, the sense of ownership is like nothing else I've ever felt. And you felt it while working for this startup? Yeah. I mean, I was sitting there with the CEO of Blaze and we were like coding for 12 hours a day, whatever needed to be done. And then like fixing up some presentation, like doing whatever you could. And if you're good at a lot of things and you tend to get bored with one thing, I think entrepreneurship is fun.

12:04Because there's always like 50 fires you have to put out. Yours is to choose which ones. Sometimes you enjoy the ones you're working on. And sometimes you don't. But that's probably true in all walks of life. And how did you come up with the idea for PolyAI? So we're one of the weird ones where like the technology is like the starting point, right? We were building conversational systems, things that sound incredibly lifelike, that understand you, even though you might have a Serbian or Spanish accent. And that's not something that even to this day exists at scale, right? When you hear an automated thing on the phone, you don't go like, oh, great.

12:39They have AI. you go like damn they don't want to talk to me right yes so that company ended up acquired by apple to make siri more conversational right but if you think about siri do you use it no right do you know someone who does and loves it my grandmother really yes okay yeah i mean it's interesting right because like different interfaces are good for uh but to be fair my grandmother would talk to a plant like she just loves talking so when no one else is talking to her she's just going to grab our phone and talk to Siri. Very nice. Very nice. Okay. So I'm not sure how much of a use case is that.

13:11Listen, that's a form of product market fit, right? Yeah. But what I realized quickly is like, people were like, hey, you know, the universal reaction to I work on Siri was serious shit, right? So I'm Serbian. We like to provoke, but we also respond in a typical way. And I'm like, yo, like, what is this about? And then I realized like, people feel it so deeply that they're not trying to rile me up. they're just like stating it like serious shit as if they said the sky is blue I'm like okay and then I'm inquisitive I'm like okay why is it shit? then people tell me what it should do is help me plan a holiday with my wife I was like that's a hard task I'm not sure that I would trust myself to ever choose something that she would universally like which is why I would just leave the whole thing to her but I'm like now you're talking about a very esoteric abstract problem that's far from people don't need a better alarm setter or anything like even if Siri was 10 times better it still wouldn't be why you choose an iPhone over a Pixel phone or vice versa in fact Google Assistant is a lot better than Siri and yet we all still have iPhones right so I was like okay this is not it this is not where we need voice agents but you know where we need them we need them for customer service because 2 % of the British population 2 % of Americans work in contact centers like the cumulative wages paid to contact center workers in the Western world are about a trillion a year.

14:38And people don't want to do these jobs. They're quitting. The attrition in a contact center can be higher than 200 % a year, which means that the average tenure of people can be as short as like four months. They join, get trained, don't like the job, leave. What about just outsourcing them to countries like India? India is getting expensive. India has moved up massively as well. so I don't think you can play the game of getting it cheaper. So cost saving and staff turnover are probably the two things that you're tackling, right? Yeah, I mean quality of service Quality of service. When people stay in those jobs for such a short time, and that can be true in offshore contact centers as well, and plus there you have the further cultural gap between the caller and the one taking the call, and maybe they're explaining what happened in Tesco and the person on the other side of the line has never been to a Tesco.

15:29They don't get what a Tesco is. They just have a list of answers to questions. And it's like, well, your order, let me see. And it's hard. It's confusing, right? So since you have basically demand, and like people are calling, and the supply is hard, and the quality is hard, you basically get to the point where as AI improves, you are able to create systems that do that job better than humans. and then over time we'll have like 90 % of all customer service happen in a fully automated way. So identify this gap and let's say that I was right there with you, I was your friend and you shared with me this idea.

16:10It sounds like a real gap but as we were discussing prior to the episode starting, a lot of tech companies they fail because they're addressing a gap that is inexistent. In this case how did you assure yourself that this was a real gap? did you test it oh man it was uh it was something right because the the assumptions going around in which year was this so we started in 2017 2017 so we built voice assistance right yeah if you went and spoke to people at that time what everyone told me was like no no voice is going away young people don't want to call they all want to chat or whatever they don't like picking up the phone and calling it's a fair argument in theory yes but you know why you don't want to call because you don't want to have a problem to begin with right now if you have a gas leak or if your internet is not working i personally don't want to call because i know that they're going to waste my time probably i'm going to be waiting way too long because the line is is overloaded you don't want to call because you're going to have a bad experience yeah what what can you think of examples where you call a company you have a good experience none probably when when i'm when i can resolve the same issue online?

17:19Yeah, so people have been pushed to online self-service and deliberately customer experience strategies. In fact, in the call centers, the automated voice is telling you, if these are the A, B, C, and D problems that you're calling for, we recommend you to go to our website and address it. Well, yeah, that's what's known as a deflection strategy, right? The simplest form. It's like, hey man, go online, bye, right? Is it to reduce the bottleneck of how many people are calling? Well, it's to try to, like, you know, many companies have spent hundreds of millions, building better websites, better tooling, because it may be a better way to serve your need to go and do it online, right?

17:54Rather than me dictating to you on the phone that my last name is Marksic, and then you go like, what the hell? And I go, Mike, Romeo, Kilo, Sierra, India, Charlie, right? And you go like, wait, was that India? Okay, okay. And by the time you say the last letter, you're like, you know what? Then you write it down, you're like, what the hell is this? This doesn't look like a name. Where are all the vowels? And then I have to do my classic, ha ha, don't worry. And they go, how do you pronounce this? They want to be nice. I'm just like, oh my God, look, I know it looks like a spelling mistake. It's hard even by Serbian standards.

18:21Can we please move on, right? And that's the choreography, right? So online I go, done, right? But it's really a matter of preference. And I think that Silicon Valley and other tech hubs, we think of it through our lens. You and I are not the average member of the population. The average member of the population very much wants to call. COVID hits, contact centers go offline, they're all still calling. Demand went through the roof. At that point, I was like... But that's emergency. Listen, the voice volumes... I'm just playing tables. Absolutely. No, no. And I believed and I agree with everything that you're saying.

18:57That was my default assumption. And it was everyone's, right? And the reason that voice AI is the hottest thing now is that the data is so mercilessly pointing the other way where companies try to push you online or to a chatbot where one person can deal asynchronously with eight people at once. they pushed you there because it was cheaper for them. Not because it was necessarily a better experience for you. They're trying to find the balance, the sweet spot in between maintaining the service quality and reducing cost. And mostly in Europe, they were optimizing around cost. Any particular industry, like perhaps telecoms?

19:29Telecoms is the worst. And airlines, followed by airlines probably. Yeah, I will say, and I've spent a lot of time in different verticals and stuff, so I can tell you about it. To their defense, the backend systems of airlines are so antiquated. and hard to change because they're interoperable with each other. The business of transforming them is really hard. There's a lot of interdependencies in between them. They're mission critical and there's these planes flying and stuff. So you can't just be like, oh, let me try a new backend for managing flights because the thing is set and agreed between countries.

20:01And there's legislation also involved. Absolutely. So it's easier for a telco to transform itself than it is for an airline. That's so mission critical and stuff. And one interesting data point that really caught me off guard is, you know that airlines cumulatively through history have lost more money than their loss making cumulatively. Yeah. In fact, I think there's a famous quote which goes, there are two happy days for an airline owner. The day that he buys it and the day that he sells it. That's it. Yep, yep, yep. During those two periods, he or she is utterly unhappy. It's a hard business.

20:40But what basically turned out is that most people love interacting. If they have to have a customer service interaction, which no one does, the default is pick up the phone and talk to someone. And you hope they pick up immediately, and you hope they know what they're talking about, and you hope they don't tell you to go online, but do the whole thing themselves. Solve my problem. Yeah. And there are good customer experiences. There are companies massively over-provisioning. We talked about high-end hospitality. Those guys over-provision. It costs them a lot of money, but it's part of the experience.

21:10And they know they have to, right? We were very big in Vegas, right? We work with like one in two casino groups there. They were like our fastest adopters, some of our earliest clients in America. You know why? Because like whenever they take a reservation over the phone for like a stay, which believe it or not, is half of all the bookings made in Vegas. whenever they do that it means you didn't go online to an online travel agency and gave away 30 % of the value of your stay so if it cost them 5, 10, 15 dollars to take the phone call if your stay is a thousand bucks, that's 300 that didn't go to Expedia or Booking.com so the cost of that, like 10, 15 dollars to service the call is nothing So it was again a cost saving No, it was a revenue making thing Revenue making, but not necessarily preserving the quality or enhancing the experience.

22:03Well, it depends on whether you look at the individual call. I think we're getting pretty close to the best of humans, right? But we are getting better every year. We weren't as good back then. So you're like a reception desk in a hotel. Someone calls you and it's like, oh, someone needs two towels there. Versus like, oh, you're calling me and you're going to book a stay for a week. Like, I want to make sure that I pick up the phone. Because if I don't, you might call someone else. so if I can take out the calls that deal with guest services people still book hotels by calling them? half and half in Vegas I was shocked as well that's where you have to realize that we are not the average members of the population this is a wild assumption it's because maybe they don't want a digital blueprint like a stamp that this happened on your booking.com maybe your wife is going to see your husband or your parents so if you're calling there's no trace maybe yeah I think a lot of people enjoy the experience of doing it yeah they do because like maybe they're driving they're like oh let me book that stay it's just like people operate at a cadence and have preferences that if you're a product guy you gotta go check what they are rather than projecting your own thing right

23:21so you have this idea walk me through the baby steps on how you put it in motion. So how do you go from idea to execution? How do you initially fundraise or did you develop yourself the initial minimum viable product? I think we were really fortunate in that we, you know, my two co-founders and me had like done PhDs at that point and were some of the first people that did deep learning for conversational systems. And our supervisor is the most famous guy in the field. So at that point, and VCs line up and throw money at you. And it was popular back then. So your pre-seed round was VC backed? We didn't even do pre-seed, we did a seed round.

24:04Seed round. And the same person who backed our previous startup just came and brought my share. Did you already have a product, a minimum viable product? We had nothing. Nothing at all. And who was your investor? Amadeus Capital and Passion Capital. So you unlocked institutional money without a minimum viable product, without a seed round just by presenting a great idea by three Cambridge graduates? Honestly, I don't even know if it was a great idea, right? The initial thing was like, hey, we know more about this stuff than anyone else. How much money did you raise? 1.7 million. And how long did that last you?

24:38We did the next round a bit more than a year after that. And what did you use those funds for? Hiring the team. So we basically pulled out about 10 people that were like former researchers into PolyAI, making it basically the biggest standalone lab for research, which is nonsensical. Research labs should not be companies. And yet you had DeepMind and you had OpenAI. So that is the model here. You first gather talent and then you work. Because we didn't have a fully functioning system running probably for the first year, year and a half. So you were able to raise 1.7 million pounds without showing a potential return on investment or a future valuation.

25:17Just an idea that we need to go and research. So go and fund my research and then we'll see how I'll pay you back. We'll see what we do, right? Yeah, because the safe space there was like, all of us had amazing job offers by tech giants. They all wanted to hire us for a lot of money. So the base case for these guys was, these guys are really good. They're going to hire a lot of other really good people. And then like - So this is where Cambridge was useful. Absolutely, right? And this is where the worst case scenario for venture capitalists is, you're an acqui-hire by Google or Meta or whatever.

25:46Do you think if you would have gone to, attended another university that was not Cambridge and you still same year, so same timings, same idea, would have gone to raise funds, you would have achieved the same result? I wouldn't have had the same idea, right? Because I latched on to the research of a guy who worked on this for 20 years. What was the valuation? Our seed round, oh, it was, what was it? It was 1.7 million round and a 3.5 pre. It's today the most diluted round I've done by a long shot. How do you come up with that valuation? so it's just an idea thinner and did VCs challenge that? oh no, I mean they gave us valuation then another one gave a higher, higher and it stopped somewhere I've discussed this so many times with friends so a lot of AI companies they even IPO completely loss making or they raise funds at a pre-revenue phase like yours without even a possible monetization model in mind absolutely how are these valuations done scarcity of talent where you know that talent not idea, talent talent, ideas worth nothing right, well sure a good idea can maybe amplify it but you're basically saying hey I have one of the few groups of people that can do this general direction idea right and then the idea changed right, by the time we were doing our series A, I had committed that we're doing context so you sit round 1.7 million pounds.

27:17Research and development, you realize that your idea wasn't good enough, and then you change your idea and you do your series A round. Well, not that even the idea wasn't good enough. It was like the idea was forming, right? The idea was forming. So we said, okay, are we building things that are going to make Siri or Alexa better? Or are we building standalone systems for real enterprises to use? And this 1.7, you said that it lasted you just a bit under two years. We always raised with enough runway, so it would have lasted us longer. But we raised more money and started hiring even more, right?

27:44And then you go to the Series A round, how much did you raise at what valuation? That was$12 million at a$42 million valuation. Two years after your seed round? A year, one year. A year after your seed round. And out of those$12 million, how much of that were budgeted as cash reserves to give you a long runway to operate with flexibility? Pretty much all of it. Pretty much all of it. Because you're operational. You already had an idea at that point, right? And the revenue model that you pitched. No. The idea I had... At what valuation, this 12 million? 42. 42. So you are valuing a company at 42 million pounds.

28:26You're raising 12 million pounds. Dollars. Dollars, so just a bit over one-fourth. And it's still at an idea stage. You don't have a product, you don't have a service, and you don't have a revenue model. We had a product. At that point, you had systems that you could call, talk, to interact with. No clients? No clients. Not a single one. A go-to-market strategy? Nope. And again, the$40 million was something that... Did you go to the market with a value in mind or you just allowed the market to determine it? I wasn't even fundraising. It just came to you? Several funds did. They competed and the valuation kept going up and up.

29:03And by the way, this is not even very high compared to what I would get now if I was doing it again. How would they go to you if you didn't have any clients? So how did they know that your company existed? it? I mean, you make a website and like at that point there was like a real, the funds that were interested in AI early on were not as numerous as they are today, but there were also a lot fewer people working on AI, right? So the value of our like IP, our models and things that we could do, you know, during my PhD, I built a thing, a system that was trained in English and, you know, you change something, how you train it and we're working like Italian, Spanish, German, Serbian, So that was like witchcraft, magic.

29:42But a lot of the funds then started realizing what's coming downstream in terms of AI. And they were like, let's get in on the action early by giving these people money. Is it not too late if someone is building a completely new tech AI startup? It's not too late. It's not too late. I think today's easier than it was back then. Today's easier to raise funds. Absolutely. Everyone and their grandmother is raising a lot of money with very happy things. Now, I think it's going to start turning a little because the market is now oversaturated with people doing that. But if you're truly different in terms of the work you've done in academia, you will still get a lot of money just based on who you are.

30:24Because they're betting on the probability that you turn that into something useful or at the end of the day that someone just buys you for who you are as well. Let's fast forward to today. How does your company generate revenue? We sell a lot of software, right? Subscription? Consumption-based. We're more like a cloud company, if you think about it. Okay. Yeah. What are the revenues of the company? 2024, latest revenues. Annualized. I can't share exact, but think like serious double-digit millions. Okay. Yeah. So. Think like growing over two times. Less than 10 million. No, no, no. Way more, way more.

31:01Double-digit, double-digit. Oh, double-digit in a million. but less than 100 it's less than 100 less than 50 can't say okay so less than 50 and the valuation is 500 yeah okay is the company profitable absolutely not you haven't really followed the CI bubble have you I did but obviously less than someone that is within the AI industry but I love the answer absolutely I love how focused you are on like real business and like well I mean like we're growing really fast because in my in my industry or my industry which is predominantly completely different than yours perhaps yours is the newest one, mine is the oldest one, the hospitality industry, right?

31:42I have investments in other industries and my background is in computer science as I was sharing with you earlier so it's really not in hospitality but it's in finance and computer science and it is fascinating for me to see that for example in my space you fight a lot to just increase 1 or 2 % EBITDA here and there and adjust the volume here, squeeze it there refine your marketing strategies and initiatives to tackle a different customer segment, so on and so forth. So it is a business. It is real business. It is like the trenches. So AI is not a real business. AI is absolutely a real business.

32:15It's just different, right? It is just the latest instance of where there is a high upfront, capital intensive R &D part that you then turn into transformational things at huge scale. I'll give you an example. We recently, with one of our clients, went live on provisioning something for the largest healthcare company in the world. and overnight we clicked a button and we started taking 50 ,000 calls a day with our system. Now the average contact center agent does about 50 calls a day. So this system click, it's there. It does the work of a thousand people. Now that is why you get the upfront investment and the bets.

32:50Maybe I amount to becoming what I am now or I'm a complete idiot and the whole thing burns because no one really wants to work with me. You didn't take any money away from me in this capital round. So early on it's really hard to do that. now it's becoming more frequent that before the large growth rounds people do secondaries and stuff we've done some more recently in the very large rounds but not early on and I think as a VC fund it would be a red flag for a kid just starting off to be like hey I would love some money so let's say how much do you own of the company? roughly 10 % so roughly you're worth$50 million what does that mean to you?

33:28well I mean how old are you? 34 34 years old worth 50 million dollars very few 34 year old people can can say that how does that make you feel honestly i don't think about it at all like it's just no no you think when you go like well i grew up this is i call it i call this show also the truth show yeah so you need to be very honest with me and with yourself no no listen i like because when you arrived you immediately complimented my watch. So you do have an appreciation for, let's call them finer things or more expensive things. I'll call your watch the ultimate CRO watch, right? Chief Revenue Officer, right?

34:10It is like you walk in and you know he has arrived, right? And look, I do enterprise sales, right? I do a lot of enterprise sales. I am a former academic who's become a really good software salesman, right? And I got to know what that is if I'm a software salesman. It'd be remiss. Yes. Because if I didn't, and I hope that you're going to sign a$5 million check, and I think you're wearing like a Seiko, you'd be like, what an idiot. Doesn't know anything about the world, right? So, you know, when you're going around Vegas selling software, if someone shows up wearing a Rishon Mill and you're like, oh, that looks like a weird watch, then you've lost the deal, man, right?

34:45So you learn things. You learn things about all sorts of things. Do you own nice things? Some nice things, but it's... Which nice things? I mean, we bought a nice house in London. That's like a real thing for the family, right? Yeah. That matters for me. And like being worth a lot, you know, I grew up in a communist skyscraper in Belgrade. I'll show you a photo later. It's a thing built in the 70s, kind of looks like the Barbican in London. And, you know, to think that now I'm like, at least on paper worth 50 million, I'm like, that's funny. That's unlikely, right? Now, money's great. I think if you're only doing it for the money, you're probably not going to be very successful.

35:25I agree with you. Because if you're really in it for the money and you're saying go to maths, by all means, please go do trading. Go do quant work for hedge funds. It is the act of transmuting your talent and skills into money. And it is a raw mapping. You should do that, right? But if you're a bit of a polymath and you want to learn about software, learn about watches, learn about whatever, then entrepreneurship is great because you're going to get to learn about all of it. Two years ago, I knew nothing about marketing. Now I know quite a bit. I care about it. If you told me two years ago that I would be passionate about branding, I would tell you you've lost your mind.

36:04Even just two years ago. If you told me seven years ago that I'd be proud of how I structure $10 million deals for companies looking to do contact-centered transformation, I wouldn't really I'd hope you were right but I don't know if I would bet on it so you get to do a lot of really fun stuff and I think that's the main perk of doing what I do like you're not really ever bored right what's the greatest lesson you've learned in fundraising in fundraising fundraising is like dating it's a game of scarcity I'm not too sure when I used to date now I'm about to to marry my my fiancé, but when I used to date, I didn't receive money, it was me spending it.

36:52I think fundraising is the opposite. Oh, I mean, right, but like the whole game of like, you know, attraction and stuff, like, are you interested, are you not, are you playing hot or cold, are you like, all of that stuff is very similar to how these funds, like, there's nothing you can do to make a venture capitalist more interested than to ignore it. For anyone starting right now in the AI space, what's the best advice when it comes to fundraising you can give them? You have to have a network. And it really depends on who you are and what you've done before. There are certain things you will want to have done.

37:32You either have done a PhD and written papers and things that have traction and got noticed, or you wrote open source code that ended up being well-known and used by others. You've shown that you can influence the whole space, right? Or you worked on real-time production models for a Google or a Meta or something like that. So you need to have a computer science or mathematics background to become an AI founder, a successful one. Is that what you would suggest? I think if you want to raise money without having a real company and a product interaction and you just want investment the way we had it, which is like, hey, we trust that you can come up with something good that will be worth a lot of money, then you need one of these things.

38:15But I think what's really interesting right now is that AI has commoditized to the point where you don't really need to have that background. Everyone now talks about the single person, billion-dollar company, and everyone's excited about it happening. I'm sure you see LinkedIn trends of people going, hey, I've gone to 5, 10 million of revenue with three, four people, and we're trying not to hire anyone, but instead we're using AI tools, it's getting a lot easier. The level of abstraction is getting easier. You can go use cursor and just write code for you. So you don't need as many engineers.

38:49You don't need to spend as much money. But you do need to believe in your idea to make it work. So I think it is shifting back to that real business. It's like, hey, if you think you can build an AI tool for that, have you built it? And why not? Why do you need money? You need some money, sure. So I think pre-seed exploded in probably the last few years. It wasn't really a category before. I've identified two different models to get a tech company going. And I want to pick your brain on which one you think is best for the AI industry. Model number one is very similar to yours. So you have an idea and you have a team of three or four different people that are subject matter experts in complementary things to yourself.

39:32And then at an idea stage you go you go and raise millions to bring that idea into a reality So that's one model. The second model is you still are a fantastic set of entrepreneurs three or four people or two people and you have this beautiful idea to build up something in the tech space and Then you look at how much what are the different things that you need to have in your pipeline in order to bring this to reality? So you look at the different business units. I need software developers. So product development. I need marketing. I need sales. I need customer support. And instead of raising funds to cover the cost of those different business units for a specific period of time, it could be one year or two years depending on your runway, instead of doing that, you sort of bootstrap it by identifying and trying to hire people in those roles that could either do it at a very low cost, at a freelance cost, so you bootstrap it yourself or you can give them equity and you get a very good software developer you give that person equity a great marketing person or salesperson you give the person equity and so on so you don't have to fundraise you don't release your your equity and also you don't really know what's working and what isn't working so you try to stay small until you identify what works and then it's when you go to the market with something solid tangible and then you turn up the volume and you put all that you throw all that money into something that can actually grow?

40:58Yeah, I mean, look, I think that at the end of the day, it's like a purely financial decision and it's about opportunity. Can you raise money or not? I think very good people in today's market don't have to work for just equity and very few would. They expect you to pay them a salary because there are other people that will give them both equity and salary, right? So it kind of becomes a must have. So either you have like personal capital you can spend on people or not. And then after that, I wouldn't really, like everyone talks about fundraising because it's It's like the first thing you kind of have to do traditionally, especially if you don't have enough money to finance maybe yourself, your salary, right?

41:31That was definitely the case for us. We were completely broke when we started this, right? As individuals? Well, yeah, I mean, none of us are wealthy, right? We come from pretty modest families. And at that point, you know, you just kind of raise some money so that you can pay people salaries. Not great salaries, but something, right? Your working capital. Yeah, and I think that at that point, fundraising, the bigger your company is and the more established, it just becomes like calories you need to grow. And there's less and less fanfare with every new funding round. And I think I'm pretty close to the point where I'm just going to let my CFO fundraise.

42:11But fundraising is fun because you basically get to retell the story of your company to a lot of smart people. So I think the way that fundraising is different from sales and why I think it's more like dating is that it's a two-way street. right they want to get the best companies and they're fighting to get you while you're fighting to get the best terms from the fund you like the most so it is a two-way thing at an early stage did you identify some critical success factors to raise funds again maybe your pitch deck shouldn't be more than 10 slides or you should dress in a particular way um yeah i mean there are many schools of thought on how you should pitch and there are many good ways to do it right it It definitely shouldn't be long.

42:51I think the vision matters most. I think one thing that Vinod Kozla, who's one of our investors, always said to me was, write out the 20 reasons why people shouldn't invest. Just put them on the title of each slide and then explain why that's not a worry. You go and you do that, and you just basically counter every objection to why you're going to be successful. Is that what you did? It's a big part of what we do before every round. It's like, hey, why is this going to fail? here's reasons one to five and here's why it's not going to happen and how we think about that. Because that's what everyone thinks the moment that you pitch.

43:27Like the upside to the best case thing, you're excited about it clearly because you're doing it. And if they're talking to you, they think it's interesting. So then the key question is, well, why not? And at that point, you need to show that you're thinking about it very deeply and thoughtfully. And you need to share the assumptions with them because you know more about it than they do and you just have to share it in a way that's easy for them to consume. Nvidia, the company, is an investor in it. Yeah. How does that make you feel when one of the biggest companies in the world becomes an investor in your company?

43:57I love Nvidia. I remember the first graphics card I had, it was a GeForce 2, MX440. And like, you know, Nvidia has gone a long, long way from just like, you know, GPUs that you put on a motherboard to play computer games. They're always important. But like, now it's like, you know, it is, the house always wins, right? At the moment they're the house, right? They're so far ahead of others in terms of just what they've built and how they've created the downstream effects. So, yeah, we love NVIDIA, and we work with them a lot. We help them because we are different from a lot of our competitors these days in that think of us as a BMW.

44:37We make our own engines. We make everything in-house, and we're very proud of it. Most others have become like vessel states of open AI, entropic. etc. Whereas for us we go down to the wire basically where it's like we like NVIDIA because they supply that wire and it's like when we work with them to create better ways to train general models and stuff then we have the way we do it on their platforms and we have masses of data that we've collected in our deployments. So we use that to create the best possible speech recognition the best possible LLM for customer service, the best possible text synthesis model for voice agents on the other side of the line.

45:20So, yeah. What is it that you enjoy the most about your job, your role as a CEO and founder of PolyAI? A lot of things. I think seeing it come into existence and have real impact on the world is what really gets me excited and what I'm really excited for for the next few years. where, you know, when we've taken a few hundred million calls already, that means that most people in the UK have spoken. And dollars, by the way. Yeah. So when you think about it like that. Yeah. I would take a hundred million calls for a hundred million dollars. Yeah. Yeah. I mean, that's a pretty good rate. Like the average cost per call is a bit lower than that when you're doing AI calls.

46:06With humans, it's higher. But, you know, you think of like hundreds of millions, that's more than the population of this country, right? Like I think probably next year we'll be at a point where every year we take more than the American population, right? Then I guess we've got to like, I don't know, set our sights on a billion calls a year, right? You look at that, that's what's approaching like a serious percentage of all phone calls happening in the world. That makes me feel pretty good, right? I think the other thing that's outside of just like the influence you have on the world, the change you're affecting in the world, that's one thing.

46:41the second thing is like the people you work with because that's everything how many people do you employ a bit more than 200 okay and you know the people you work with i'm a big subscriber so talking about nvidia i love jensen huang i love his honesty i don't know if you've seen like his interviews where the guys asked hey like would you do this again would you guys like nope are you enjoying it yeah and no and i'm like oh my god finally someone honest because you know you You just get all of these run rate Californian zombified people. We're like, oh my God, yeah, I love to change. Okay, cool.

47:14You were written those bullet points before the interview, right? Whereas you just see Jensen in his leather jacket going, why are you successful? Because they work nonstop, literally nonstop. I was like, cool, okay. It's good. It tells people what it takes to succeed, right? What is it that you don't like? Well, the other thing that's really important is do you like the people? So what I really love about Jensen's approach is, like have 50 direct reports and expect things of people and grow people and be open and create a culture where you can tell someone why they're wrong without them kind of like scurrying away feeling humiliated right because if you create that culture you create network effects you kill the bloat of middle management and you create like a thing that is a pleasure to work on because it attracts the right kind of people and repels the political you know little empire builders within companies.

48:03So I enjoy when I see people growing, right? My favorite stories of our people are people who joined this post-university and, you know, like now run teams of like 50 people. Like, that's awesome. That's real. That's real growth, right? Their trajectory is the trajectory of this company. And, you know, they'll probably go on to do things greater than I've done. And that's like what gets me excited. What is it that you don't enjoy? Because not everything is a walk in the park and it's impossible to like everything. about anything. For sure. I mean, everything I like, I partially dislike. It probably just depends on how it's going at a particular time.

48:41I don't really love management as a discipline. I've noticed that there are people who are really good managers. They enjoy it. Give me an example of something that a manager would do that you don't like. Right performance reviews. Right development plans. All of the bureaucratic stuff that you didn't want to do so you started a company instead. and then as you scale and your company becomes a real company, you realize that all of these things are needed. Whereas really what you really would much prefer to do is be on the front lines. And a friend of mine would say, wait, hold on, do you want to be a warlord or do you want to be the king?

49:19And those are two very different things because many CEOs are warlords. They go and they sell deals. They look at key implementations. They focus on product things that really matter. And they're really good at it And they continue to do it despite all the administrative burden, and that's what makes them great. Steve Jobs is a famous example of that, where the guy was probably the chief UX guy for Apple all the way through. Would you consider ever stepping down as CEO? Because I believe that a great founder doesn't necessarily mean that he or she can be a great CEO and vice versa. So for example, I think at different stages of a company's maturity, it requires a different type of leadership.

50:00I believe in the early stages, a founder profile style of CEO, very passionate, very driven. It's about strategy and the vision. It works best. It works best if you can do both. I think so, but I think it's not exactly, it's not necessarily. So for example, in our case, so the revenues of our company are also to teach it in the millions. And we are right now in a situation in which, aside from what I enjoy doing more or less, I can clearly see what's very much needed. and a pain point for the company. And then I ask myself, am I the best person right now or one of the best ones in the market to satisfy that pain point?

50:39And clearly, frankly speaking, I am not. I think I am a great founder. I think I have the vision. I'm very good. I'd like to think that I'm good at fundraising and strategizing what's the next step for our company and so on. But I think that there are other people that have done this for groups that are perhaps even bigger than mine that can do a better job because they've done exactly that. But will they care as much as you? But that's different, right? That's why I'm still not stepping down as a founder and also as a majority shareholder. And the CEO would effectively report to myself in the board.

51:14But someone that has these A, B, Cs, and Ds has the methodology and doesn't need to do a lot of A, B testing and still making errors at very high levels because making errors when your company is under 10 million revenues is something that you can pretty much work with. But anything on top of 10 or 15 million revenues, the cost is greater. So I feel like finding a balance in which you can have a C-suite that has done this before and has more experience in those particular things, but they still report to you in one way, at least for my industry, I think is quite effective and efficient, but it's also not easy to let go because as you said, that's my first initial instinct.

51:58who's going to care more than me. Why would you let go, right? I think there are many different kinds of CEOs. And there are, I think, the things that you will have to learn, right? You can't be financially illiterate. You can't be, you know, like you have to be good at very many of the things so that people reporting to you have respect for you at the end of the day, right? Because people don't want to work for people who are like worse than them, right? They can be better at you at what they do, that they're subject matter expertise, but you have to know enough to be dangerous, right? I think that's the fun part of being the CEO.

52:27those mistakes and stuff that you make will be countered and it depends on the nature of the business right it really like i can't imagine ever stepping down because i'm committed to growing in a way that is needed for this company to succeed in its mission which is no one has done this before and i care more than anyone else does so i think that that's a very interesting point because i very much agree with what you're saying no one has done what you're doing before what poly AI is doing before. In my case, in the hospitality industry, there is a blueprint on how to scale it. So for example, how do you go from 30 million to 200 million revenues?

53:05So there is a blueprint and I'm pretty sure that there are people that have done that in their previous experiences. If you look at the example of Jim Shark, Ben Francis, he also stepped down for a period from the CEO role and he onboarded someone that has done something relatively similar and that worked for them. And I think what you're saying is another great example of this because he's building a clothing brand and it's not the first clothing brand in history. In your case, you're still doing something that is a sort of test question mark. So I can understand that. You look at the great companies of tech companies.

53:38Google brought in Eric Schmidt and that worked. Zuck scaled. Did he bring in seasoned professionals like Sandberg to be the CEO? Zuck acquired. Yeah, and really well. He's like the M &A CEO of our time. Right? Phenomenal. Yeah, I agree. Like it's, you know, for all the critics of him, because he's not like maybe polished. But there's a difference in between inorganic growth and organic growth. Sure, but I mean, he first organically built an empire as well, right? So then he was smart enough to not have the vanity to believe that he shouldn't acquire it. I'm not disputing that. Yeah, I agree. But, you know, he's an example of someone who brought in the right expertise at sea level and yet continued to do his job, right?

54:21So in his position, he can place the bets. He bet on AR, hasn't panned out yet. Doesn't mean it won't. But while he has that seat, he can re-engage and do it because he can. And that's good. It might one day turn into the biggest thing and make Meta 10 times the size of NVIDIA and Apple and others if he makes the right bets. And then him at the top, you believe that the DNA and lifeblood of Meta is that kind of company is still there, right? Waiting to pounce. Similarly for NVIDIA, right? But you also have other great companies where the DNA moved through a different CEO, right? Like you look at Satya and he's arguably the greatest tech CEO over time, right?

55:04Just as like the professional CEO. And I don't think he at all feels like he's not Microsoft through and through because he is. He's one of those guys that joined early and just like became the greatest, right? So you're right. I don't think there's a blueprint. also I think the face of the company doesn't necessarily need to be a CEO level a lot of other companies particularly consumer brand companies the founder is the face of the company but he or she is not necessarily the CEO you can still feel like the fabric of the company is very much correlated with this person's DNA but this person is not running the day to day show.

55:43Yeah I mean look I don't think Zuck is running stand up meetings asking each P &L owner how last week went. I've had this discussion multiple times with friends over dinner. And there's a divisive opinion on how the AI industry will evolve. So I have a set of friends that believe that... And ultimately, it's going to be chopped into three or it's going to be distributed to three very big companies. And those three companies are just going to acquire every middle and small company and ultimately is going to be leading the AI space. Then I have another set of friends that believe that the AI sector is so big that it's going to be fragmented and owned by multiple players.

56:32I can imagine what you hope for, but what do you think? I think the power of cloud is nothing to scoff at, right? If you read like Yanis Forofakis... Can you simplify that for me? Yeah, so if you read Yanis Svartoufakis, the former Greek finance minister, he calls it techno-feudalism. He's basically like we're in a new post-capitalist world where it's no longer capitalism. It's like fiefdoms where anything you build on top of technology pays a tax to Google, Microsoft, or Amazon. Because you're hosting everything on their land, effectively. They own the compute. And then you have the three layers of software.

57:09You have infra, you have platform, and you have application. right? And the infla layer of course Nvidia is like another player so let's say 4 right? But between them they're very poised to own that layer of AI right? Then you kind of go into that like foundational model. It's very similar to for example here in the UK that most of the land is owned by the crown state then some of the buildings on top of that land are owned by different land arts and then you have tenants inside so that's okay. Yeah and you know like I think that they're in a position to be very successful companies for a long time.

57:38In this case Amazon and Google are and Microsoft are the crown state. Yeah, they are. And you build software and they provide you a service and computer and everything. You're no longer stacking racks of servers to do these things. So you will definitely pay them. That's a good analogy. It's a new form of real estate. It's so true. Yeah. And what they're trying to do is not be the crown estate where you pay like, what's it called? If you pay every year, like the... Council tax. Council tax is through government. but there is like a different like um if you're a leaseholder you pay like some you pay you pay the business rates or you pay the service charge yeah yeah no no i mean there's like a certain like small amount that you would pay to like the the like the final owner of the land on which your like flatter house is built right unless it's like that's the leasehold freehold it's like the weirdest thing in real estate maybe we don't go into too much detail because anyone listening who's not in the uk will have no clue what we're talking about right um and there's reform coming anyways probably.

58:40But the interesting bit is they are there and they control everything that you build on. But the crown doesn't own every company in the UK. Maybe they get a cut of some real estate profits, but really it's no longer that meaningful or relevant. So I think it really depends what we build on top. Now, in terms of controlling distribution of software to companies, the cloud is very powerful as well. So today you'll have a lot of you know it becomes very political it's kind of like if you're like a small country it's like where are you buying your guns from are you part of nato and buying it from the americans and brits and the french or are you buying it from the russians if you're buying it from the russians why are you buying it from the russians why are you not buying it from nato are you buying it from the chinese right so much like the orwell's kind of like you know three like power blocks you kind of have that happening in software now where it's like oh you're an aws shop well like we prefer doing business with Microsoft.

59:35They're like soft influencers in the flows of revenue in the upper layers. Will AI become owned by them? No. Will they be a big part of AI and distributing it and owning it, especially when it comes to foundational models and how you use and consume them? Yeah, I think they will. I don't think that there's a long-term business model where an open AI or Anthropic, they already are aligned powers. They're like the UK to the US when it comes to Anthropic to AWS or opening it to Microsoft. And that's now a rivalry where opening it is trying to break away, but can't they really? And then they're kind of competing, but they're also allied.

1:00:13So it's interesting, right? It's very political. So what you're saying is these three companies, Microsoft, Google, and Amazon would own the underlining... They will collect rent. Exactly. Now the question is, what will you build on top and how will you deliver value? Can you long-term do it in isolation from these large platforms? Because they provide a lot of software. If you're a company, you're probably a large company with thousands of employees. Because ultimately AI is just another SaaS, software as a service. It's not necessarily SaaS, right? That's where I think there's a lot of discussion of you see people saying SaaS is dead or there's no future for SaaS because all of this can just be rebuilt easily.

1:00:55Unless you have network effects, there's no future for SaaS. I don't know that that's true. I don't think it is. I think that when you buy enterprise software, you're kind of buying a throat to choke. Like if I'm paying you millions a year for this thing to do something to my organization, if it stops doing that for my org, I get to call you and say, hey, like this doesn't work. Are you going to fix it? Or am I not going to be paying you this in two years' time? And that's really what you're paying for. That's the implicit contract. And you hope that you and others are asking for the same thing and they will do it.

1:01:28right um so i think that ai is not really its own like thing it's just the next enabler of new kinds of technology right so some of it will be sas and licenses a lot of ai is just doing work so it's really units of work right we charge per minutes mostly sometimes we price it in an outcome based way where it's like if our system manages to handle the cold you pay and if we don't you don't right but that's similar to how like a contact center company selling labor to people might do it right they might say you pay me you know 15 bucks per hour per effective hour or like pay me 3 million a year and i'm gonna handle your calls and i promise to pick up 90 of them within one minute so that's just business on top and how i commit to providing a service for you we try different ways we compete with each other among providers of any service and you see what that ends up working in the real world, right?

1:02:24So is it SaaS or is it like more of a cloud model? And that's another reason why cloud is interesting, right? Because a lot of the intermediary companies that have built software for these things charge on a kind of like per seed basis, right? For humans using it. Now, if you have fewer humans using it, that kills their revenue base in a way that is pretty bad. So for example, if you're a CCAS, contact center as a service company, you sell software that contact center agents use, right? so if someone sells AI you have fewer seats and that really hurts their revenues so they're not in a great spot whereas if you're Google or Microsoft or Amazon you might have your own solution for that and you might lose some revenue on that solution but because people are hosting their AI on your server you make more money that way so in that way they're in a really good position to always win and that's not true for a lot of SaaS companies that currently exist because their revenue base is eroding because there might be fewer people using their software.

1:03:26I want to talk about your co-founders. In entrepreneurship, the relationship between co-founders can make or break a startup. In fact, it's the second reason why startups end up failing. You are one of three co-founders. How would you describe your relationship with them? Deep friendship, love, and respect. a co-founding relationship and I've said this to a lot of friends that are thinking about starting companies it's like between however many of you are starting it you don't have all the skills you need I think the key thing is to deeply respect and believe that your co-founders are way better than you at some things and do they have a different set of skills than you?

1:04:10yes yes, very so your background is in mathematics and computer science? On paper, we all looked kind of the same. So how did you split your responsibilities when you co-founded the company? It was very simple. I'm like the talker. I love talking. I love doing that part, right? So like me doing this job made immediate sense, right? Where it was just, I wanted to go and speak to investors and to clients and to do all that, right? And while both of my co-founders are really good in those conversations as well, I think they'd find it a slog more than like a cool thing to do for days on end. Whereas, you know, we go through a funding round, I leave energized, even though I might've like not slept for weeks, right?

1:04:53But like, I enjoyed it. I've retold the story of a company. I got us something that means a lot to me, right? And like, of course they care for it and they contribute and they do a lot of work on it. But you know, like I think - You all have the same equity. Yeah. And from day one, you suggested that you wanted to be the CEO. Yeah, I think that was... And was it your idea? Was it a combined idea? It was a combined idea. I mean, like, you know, we talked about it. We had discussed it for a while and decided that, you know, hey, we really enjoy working together. We're going to do it. And, yeah, we're all equal co-founders.

1:05:30And then, you know, you kind of figure out who's going to do what, right? And then, you know, roles evolve and all that. And, you know, by the time we're at this point, it's a professional company, right? so one of my co-founders runs product and tech and the other runs engineering for all of our deployments and deployments are what's really important here because AI is actually not SaaS our kind of AI is not SaaS it's actually more like on-prem software where for each client we need to build a very specific thing like it needs to do deep workflows and integrations but do you deploy it under infrastructure?

1:06:03for the most part no for the most part it's like multi-tenanted cloud deployment but you know you kind of like co-founders wise you're just kind of like there's a lot to do there's really a shortage of things to do right and over time you hire more people and the team evolves right and responsibilities evolve both according to capability and interest right would you advise anyone to start a business with a friend

1:06:31i would advise that people try to not hire their friends that's like power asymmetry that's I think you don't need in your life unless you really need it in your life co-founding is fine but know that it has risks right? Did you make the mistake of hiring a friend? No no thankfully all of my friends that we hired were like with us for a long time and had a great time right but every time I sensed like trepidation and huge responsibility on not failing them, right? And that not impacting our relationship. Co-founding is a position that is equal, and that's, I think, a lot healthier. So co-founding a company with a friend is good.

1:07:17But also, there are a lot of people who maybe want to start something and have a strong idea and then feel compelled to have co-founders. And then if they just found a friend who is in it because it sounds interesting, that's a really bad idea, right? It should be like they both care and have something very meaningful to contribute. right? If that's not the case, then you're just mucking about, you should not be doing that. Like, treat it very seriously. I think, you know, it's almost become the new thing to do to start a company. You know, you hear like all the finance people of London are like, oh yeah, I do finance.

1:07:47And you know, you know that they enjoy their life. And you know, it's probably the best ROI on efforts of anything that you can do in London, right? Steady, large paycheck and stuff. And almost universally, you hear a lot of them going, but you know, I'm sure I'll start something on my own one day. I'm like, why? Why would you do that? channel Jensen? Why would you? I fundamentally agree with you. I think that particularly movies, social media, they glorify it. They're telling us that the highest purpose-driven career is the one of an entrepreneur. And I fundamentally disagree with that. They tell us that also the highest level of success and respect only comes from entrepreneurship.

1:08:26And that's how you should define your success. I think success is very much related to happiness. And I think the overall aim in our journey is to identify what truly makes us happy. So for example, I've discussed this in previous episodes, but my girlfriend, I think she would be a good entrepreneur. I'm still not 100 % sure if she would be a great entrepreneur, not because of her skills, but because she's a person that thrives when things are foreseeable. I'm the opposite. when my life, particularly from a professional perspective, is more complex, the better version of myself I am. So I am sharper, faster.

1:09:08I also energize from conversations like yourself. And I don't think that makes me better, not worse. That just makes me me. And by getting to know myself, then I can feed myself what makes me happy. Absolutely. Right, if you turn stress into energy, you should be an entrepreneur. Well, you can be an entrepreneur. You'll probably be good at it. and if you're good at it, you're going to enjoy it and then you might be happy. It's a precondition, not a sufficient thing. I think my wife would be a killer at being a CEO of a company. But does she want to do it? No. I think she's very self-aware. It's like, okay, that's not the right thing for me.

1:09:45Great. When people glorify founders and stuff, why? Be a top surgeon. Be a top surgeon. You will aid countless people it's so non-controversial you'll be paid a lot of money as well you're gonna have a great well i don't know if you're gonna have a great life it's gonna be a tiring life and stressful but if that gets you going like you will have contributed more to the world than the vast majority of founders let's look into your personal just for a bit so before we started the episode you mentioned that you have two kids how do you how do you do this while running an almost a billion dollar empire it's hard um like young kids yeah isabella is two adriana is almost one um it's been the happiest thing that's ever happened to me having them right because they just like act as this like endless bundles of joy and fun and every moment with them is like like it just you know fills you with like a stabilizing happiness that allows you to be better at work and stuff.

1:10:52It's not easy. You like that stabilizing happiness. Yeah, yeah. I, you know, a friend of mine, he's a venture capitalist and he was like, I realized how much time I was wasting like helping on stupid things when I got kids because he became a lot more ruthless. I said like, I could stay here all night helping you figure that out or you could just do your job. Like, and, you know, when the trade-off is like, you have one more hour and like you're going to see your kid before they go to sleep or not, like then like, you'll start making decisions that are not always pro-work. What would you advise a parent that is building a company, but also is parenting?

1:11:30It's really hard, right? Like I think that, and I say this to a lot of people who think of starting companies. Think about the right place and time in your life, right? Like, do you have a partner or not? What are your plans? Are you planning to have kids? Does your partner, your wife works? She works, yeah. But I think like European maternity leaves of one year are really good, right? When I compare that to a lot of my American friends and colleagues, it's really tough. That's much tougher, right?

1:12:00You need to set it up. It's expensive as well to get the help and everything you need. So are you set up to do it? You shouldn't start a company unless you have a very clear plan or maybe a lot of savings. If you're about to have your first kid, for pity's sake, timing is everything in life with everything, right? So sort your life out the right way, right? And I think, you know, fundamentally, you're best off starting a company before you have kids because then you can be way more committed to the company and it takes a lot, right? I think when you're in the growth stage, then, you know, you slowly start turning into that CEO you were describing.

1:12:38Someone else could do a lot of my job, right? And that then means that, you know, it's less existential. The bigger the company is, it's always existential, but it is less existential on a daily level, right? Where you can then force yourself to organize the company, workflows, hire the right people to have a semblance of a life. I'm not sure I do, if I'm honest with you. I'm not sure I've done this right. But your personal life changes in that. Well, I mean, it's your choice to make, but for me, like all of my time that's not work time, which is most of my time, is family time. And the rest of my social life basically doesn't exist.

1:13:17But I'm very happy, so it's okay. But I think people are different. I don't know that this would work for everyone. I think it's about finding the right balance. Because ultimately, the biggest resource that we have is time. And time management, I think, is probably the most important skill that we need to learn to master. And how to balance it in between your company and your family time. It is fundamental. And perhaps maybe in some industries that balance is different because some of them are more physically demanding and you just need to be there and you cannot outsource or delegate as much.

1:13:54It really depends. But I think there's a balance that anyone can reach. It's a hard thing that I think people pursue and I think most founders are pretty bad at this, right? Because they just do more and more of you anywhere can probably help, right? So it's hard. I don't think I'm very good at it, if I'm honest with you. I could probably work less and achieve. I don't personally believe that more of you always is better. Not necessarily because it's of you, but your energy is finite. And I fundamentally believe that sleep is very important to be the best version of yourself. And if you're not getting the right hours of sleep or eating properly or exercising, the work that you're going to be doing is going to be far less meaningful, impactful and accurate.

1:14:42so you're probably going to be making more emotional assessments but there are many shades of that I mean you can see from the bags under my eyes that I don't really do it all that well and the truth is do you think you would be able to perform better if you exercised a bit more and sleep better definitely why don't you do it because I think there are trade-offs to make different points like right now the balance that I'm striking is better for what I need to achieve. I agree with you. You can't really do it for like 20 years. You can do a few years of extremely hard work. I've heard someone say this where it's like, you look at tremendous successes.

1:15:23They've had a few years of extreme disbalance where they went all in on what they're doing. I'm not saying it's necessary. You can burn out. You can burn out easily. And you burn out and then you get back and you feel it. Not only burn out, but I had an amazing conversation with a stress management expert. and I've had a similar conversation with a neurologist. And the common denominator in how stress affects our brain is that parts of our neocortex and our frontal lobe, which is very much related to our memory and our cognitive functions, gets very much affected by big stressors and by extended periods of stress.

1:16:01Yeah, yeah, yeah. Because it's an addiction, right? Stress, you develop an addiction on it and then... Not only that, but so your stress increases several different hormones. One of them is your cortisol levels. And the problem is that your cortisol levels remain quite high if you're constantly under stress. And you are changing the baseline of your cortisol levels, let's say, from medium to high. And that has a profound multilayered impact on various other organs of your body. And that can become quite dangerous. So you can not only lose your hair, but you can have heart problems. you can have several different impacts that it's very hard to then reverse engineer or reverse engineer for that matter.

1:16:45So my humble and personal advice to you would be, which I think you're doing great things and I think you should be doing it half an hour a day that you take it for yourself, you exercise a little bit and that you prioritize to sleep. I think it's going to do wonders to yourself as a person, to how you are with your family, but also how you are as a CEO. No, for sure, for sure. And this is probably the biggest lesson learned that I've learned because I think we live in a culture in which both in Europe and in the US, for example, in modern societies in which when someone is asking you, how are you?

1:17:19The typical answer is, all right. I'm overloaded, overworked, stressed, surviving. And it almost feels like if you say, I'm fine, you're almost saying to someone, I'm not working enough. yeah right and pardon my french but that's bullshit yeah that is bullshit it's a choreography it makes no sense exactly so i very much um nurture and advocate for a healthy long-term sustainable lifestyle because if you still want to make that billion dollar company or that hundred million or become the best father or the best whatever it is the best version it's a marathon yeah it's a marathon exactly so consistency is very important and maintaining a balanced life that is sustainable I think it's vital it's paramount and it's definitely going to increase the chances of you achieving your targets yeah yeah because it's a marathon you have to last yeah yeah what does Sam Alton say it's like you just need to not quit and the best way to do it is to not collapse on one front exactly so for example in this show as we were discussing earlier we exclusively interviewed in season one Forbes Under 30 cross sector so some of them like yourself coming from the tech environment some others finance consumer goods and so on and there are a lot of different lessons learned but there if there is one which i would identify as the strongest correlation in between all these very successful entrepreneurs cross industry is that they didn't quit yep now in order for you not to quit there are multiple things that you need to have under your backpack so one of them we've discussed for For example, if you are mainly looking to make money from this industry, you're going to be quitting very quickly because you're not going to make money overnight.

1:19:06But another very important one is exactly what we're discussing, trying to avoid burnout as much as possible. So you should enjoy what you're doing. And in your case, I clearly can see that you enjoy what you're doing. And that is going to prevent you from quitting in the days in which every fiber of your being is telling you quit. Because that happens in entrepreneurship, right? You're building something new. is quite volatile. It's a living, breathing thing. And that's going to impact your emotions. So I think a very good way of also supporting yourself during those periods. And this is something that I do myself.

1:19:40And at the end of this year, we're going to be employing over 400 people. Maybe the valuation of the company is not as high as an AI company, which I would love to have those valuation methods. but from a workload perspective, from a workload perspective is very intense. Absolutely. Hospitality is like the toughest industry. And I found that one of the strongest, one of the strongest and most healthy things that I can do for myself to be able to do that in the most successful way is taking care of my sleep, taking care of my diet and taking care of and exercising regularly. And that a lot of people believe, and sorry that I'm extending too much on this, but I believe it's very important for any aspiring entrepreneur that is listening to this, a lot of people believe that it takes so much of yourself.

1:20:26Oh my God, I don't have that much time because I'm building my company. And sorry, and again, pardon my French, but that's bullshit. It's bullshit. It is bullshit. I've done it. And I'm telling you, not only I'm running what I consider right now a large organization, maybe for some others it's small and for some others it's huge. But for me, I think it's becoming quite a serious, stable type of company. But also I'm running a tremendous podcast with a team that is amazing, I'm writing a book. I discuss this with you. I'm entering the AI space. Very good. And then people might ask me, how the hell do you find time to exercise or eat well?

1:21:02You can. Yeah. You can. And anything else is an excuse. It's an excuse. Yeah, yeah, yeah. I mean, you know, I think that like, especially if you're like a B2B software CEO, the place where that gets tough practically is just like you're on the road a lot, right? And a lot of people who travel a lot just get a bit unhealthy because it's like everything you know is like, you know, the enemy of like a good regime is just like not having a stable, continuous thing. You don't live in one place. You're like in a different hotel room for like three weeks straight every night. At that point, you know, like client dinner where you scoff down like a giant steak is like the only joy you have in life for days on end.

1:21:41That's just like practically it is what it is, right? But yeah, you kind of have to like... But let's look at that lifestyle. So you're on those four trips. you're going for dinners from one hotel to another. Let's assume that you don't have the luxury of going to expensive hotels where they have a gym. Right now, what could you be doing with your day if you have half an hour to exercise a little bit? No, you can totally do it. You can totally do it. It's doable. It just takes like a commitment. Determination. It takes commitment, yeah. Exactly. Yeah, for sure. Now, going back to your journey in building PolyAI, what's the biggest challenge you faced?

1:22:18well I think the biggest transformation that I had to go through was not believing that we just have great technology and things will work out until like alright kid how are you going to make this actually get deployed at scale right what I didn't know when we started and when we picked contact centers is that and that's actually one of the reasons we're a great company now it's because we're the best at go to market and that didn't come overnight but it's basically how do you convince companies to change that 1-800 number they say in the US like the default line where people call their business because chatbots on the website that's cutesy, someone might use it, someone might not whereas this thing is like everyone having any reason to contact you or escalate an issue calls that number, so I call it open heart surgery so convincing people to try out something new there that's like a pretty heavy maneuver I only now have the appreciation for how difficult it was in that everyone else is trying to oh let's make your contact center agents a bit better and then over time maybe we can put AI in front of the customer we'll put a chatbot on your website see how that goes and then maybe we'll move into voice somehow whereas we just kind of like took on the hardest challenge which is like fully automated voice on your main phone line which industries were the earliest adopters hospitality was huge for us.

1:23:43Logistics, retail, financial services came then and are very strong. And now healthcare is growing the fastest. And it's just like, that's very heterogeneous. So like, there's a point where a lot of people tell technical founders, and I found that in Europe, they do this a lot and it's, quote you, complete bullshit, right? They go like, you just now need a great sales leader and they can sell and you do what you do, be a cutesy little academic, working maybe on product and stay in your lane. The answer is if you can't swim in that other lane, you have no business of being the CEO of, say, a B2B company.

1:24:16If you can't communicate to the world why your software is going to be huge for them, then your software is probably not great for anyone. You probably don't even have software. You just have disparate technologies. So that person might as well give up on that CEO journey immediately. And in Europe, especially to technical founders, the advice of, oh, nerds, stay in your corner and we'll hire a great salesperson. So you're really hiring, because that's not a salesperson. A salesperson sells a fully defined thing. What you need is a problem solver who can continuously feed the information back to the product and evolve the company.

1:24:49That's your job. So I think for us, the challenging bit was that transformation. And it was also the best part when I look back. There was a year where me and Brad, who's our best sales guy, he's technical and background as well. There was just a year where we were like, okay, we're going to stop trying to hire salespeople and stuff. We're going to take all the sales calls between the two of us often together. And I look back at that and my calendar was like, blue means sales meetings. And there were like four or five a day. I've never done more work faster and better than that. Because it's challenging.

1:25:26You're like, oh, I need the answer to that question. I'm going to bullshit my way through it. Then quickly, I'm going to go back and figure out what I actually said. Can we do it? And that stuff is the switch for product market fit. for a software company. And the answer is like, it's all you, buddy. Like, there's no salesperson that's going to help you here because this is not sales. This is core, like, search for product market fit. So go get out there. And if you can't, well, give up now. I love it. Yeah. You mentioned something that I find very interesting and we can deep dive into it. You mentioned in Europe as a market and then obviously there's the US, right?

1:26:04Probably these are the two right now dominant markets for emerging startups. I've discussed this also several times with friends, particularly tech entrepreneurs. Most of my friends feel like if you're finding, building a tech startup in Europe is more challenging than in the US, mainly because the VC ecosystem in the US is more supportive of early stage tech companies. For example, like yourself, your company. Even though you did it in Europe, I want to pick your brain in this. My friends feel like if you are raising funds in the US for an early stage tech company, even if you are at a pre-revenue phase or you don't have a revenue model in place or even defined, they're still going to believe in your idea.

1:26:52That's why the US is probably right now leading the way. But in Europe, they usually are more conservative, more risk adverse, and they want you to. So you agree with this? Well, look, I mean, I think there are some great funds in Europe, but for the most part, they have much lower conviction, right? And it's changing. It really is, right? Especially like the UK and most of the UK. So you would say that the UK within Europe is the best market to... Israel is better. Israel is better. If you think of it as like greater Europe, Israel is better. Per capita is vastly superior. To raise funds? it has a better entrepreneurial ecosystem people have done more, achieved more there's like role models to look up to what I know about Israel when it comes to tech startups is that they're very good in cyber security for example they're phenomenal, right but they're very good at many other things they've built so many good companies but the difference is a lot of the old school European venture capitalists are like show me a revenue plan if you're truly stand out At the seed stage or at what stage did they ask you that?

1:28:00Every stage they were asking about it. They were like, five-year plan. I was like, are you for real? If you want, let's take the best restaurant operator who cares and knows. They're not, well, I mean, they're probably pretty good at telling you how the business will evolve. But you take an AI researcher. They're going to have to learn a lot and hire a lot of good people. They can't answer that question. It's almost a red flag if they can. Because why are they doing that? It's not that clear. There's stuff to be figured out that will fundamentally make the business so different in five years that this plan will be bullshit.

1:28:33And yet they all play this choreography of how will you get to profitability? The answer is it doesn't matter because there will be funding if you're growing violently fast. There will be funding and the right thing to do is to continue to grow violently fast to capture as much of the market. So why are we talking about it? Why are we wasting our time, your time? So what I especially found in those like, you know, 2018, 19, et cetera, Europeans were slow. So slow. Like, oh, yeah, we can meet quickly. How about in two weeks' time? Like, I'll be dead in two weeks, right? What do you mean tomorrow?

1:29:05I agree with you. Americans were, you know, I'll tell you how our like Series A played out. We had a term sheet. So before your Series A, you see the investment round, those venture capitals? They're British funds. They're British funds. Okay. But from then onwards? Americans. American. Well, our last round actually was co-led by NVIDIA and Hedosofia, which is headquartered in London, but it's very much a global fund. But they're very American-style. They're fast as hell. And I think there are people in Europe who are bringing the playbook of U.S. funds, and it's changing slowly, right? If you were to start all over, what would you do differently?

1:29:38Move to America immediately. Start it in America. Where? It doesn't matter? California. Still California? Bay Area. Bay Area, yeah, yeah, yeah, 100%. I'm probably there no more than a year full time. Anyway, right? And I'm there all the time. It is the center of the tech universe. The network affects. And a bit of arrogance. Like they don't really think the world outside of the Bay Area exists. It's all cutesy, right? And outside of America, for sure. Like London is... Where are the headquarters of your company? We have three offices. We don't really have a designated headquarters, right? So we've got offices in San Francisco, in New York, and London.

1:30:14London is the biggest still. but most of our go-to market is in the US. We've got C-levels in each of the offices. We're quite distributed as a company. Your co-founders are based here? Yeah. The three of you technically are based here. Yeah. I mean, I'm in America half the time. Okay. And that market is just superior, right? It's not that Europeans are, yeah, we have less experience. We have fewer people with playbooks where they've done marketing or sales or et cetera at this scale, right? So it's just harder to, you know, when you talked about like, do you grow to become that or do you hire in the right talent yeah it's hard to hire the right talent in europe when almost no companies have scaled to like a billion dollars of revenue in software right yeah there's like probably like well i don't know how many very few probably like under five european companies that did that whereas in california it's run right you walk down the street in palo alto and you're like ah like you know series d marketing guy yeah series c cro right and it's just like you almost pick any of them and they would know what to do in terms of the playbook for their department so um it is very strong right nicola you're and the list in forbes 30 and their 30 list again if you're very honest with yourself how important was this for you not really i think that there was a period where like people were really gunning for it and it's definitely a big deal, right?

1:31:41And it definitely helps people and there are people who care about it a lot. Did it help you? Because you told me that Cambridge, that credential helped you. Did Forbes Under 30 help you raise funds? I can't tell you for sure. I don't think it was essential, but it's definitely a good thing and it's very helpful. There are buyers, there's definitely help in deals where they treat you like a minor celebrity. It's a great award because it gives people the reassurance that you've already made it. And that's a big deal, especially when you're a startup. Anything that can give you that aura of validation, this is a safe bet, this is a super strong individual, a super strong company, is definitely helpful.

1:32:23So I think in sales, it's been helpful for sure. I know people who are dying to get it and who thought of it as a key milestone before they're 30. I mean, I got it. It was cool. cool like uh it was cool it's not no doubt it's very cool like um but i don't think i'm over obsessed about it at all nicola how would you define success uh it's my wife's favorite question how do we define success um depends on the situation right what kind of success in business or in it's a great question well you said you said being happy right i would not have gone there right you know no one has asked that question back and?

1:33:05they've assumed a definition of success some people answer it from a personal lens some others from a professional some people it's a hybrid of both how would you define success? for me? yeah

1:33:21I love prevailing against the odds right the whole like David versus Goliath fight and warrior I think it's a very Serbian thing you know like we have this attitude of it's us against the world I think it's an Eccola thing Yeah, I mean, it's, I enjoy the fight. The fight that is asymmetric, where you should lose by default, and you have to find a way to win. Do you think entrepreneurs are made or born?

1:33:48It's the whole nature versus nature. I think it's a bit of both, probably. Like, I used to be, like, fully in the made camp. I didn't believe that there was such a thing as, like, talent or aptitude. I thought it was just hard work. Yeah. The longer I've run this company, I have come to believe that that is not the case. And that nature is a prerequisite. You need both, right? But it's that talent that gives you that initial edge in something that then gets you to work harder than others. And then it becomes a virtual cycle. Not everyone can be great at everything. So for entrepreneurs, I don't know.

1:34:24According to your definition of success, would you say you're successful?

1:34:34I'm alright I am perpetually in want of more and I am I successful? Sure, by traditional definitions. Do I think I've become the best version of myself and that relative to both effort and opportunities have achieved the most I could have? Probably not Thank you in so many ways for being so honest, direct, and for sharing so many insights about your company and your industry that a lot of people don't really know. It's a hot topic that everyone is discussing, but few people actually really know what's going on. You fall under a very particular type of category of entrepreneurs, especially entrepreneurs in which right now young university students look up to.

1:35:34I recently delivered a lecture at Bath University and it was very interesting to see that most of these students wanted to become entrepreneurs in the artificial intelligence arena. And a lot of them don't really know what that even is. Not all of them were computer scientists. Other people were business majors. but they still wanted to do something in AI. After having chatted with you, I am profoundly happy to see that I met Nicola, not only the CEO and the founder, but I met Nicola the person, Nicola the human. And I very much hope so that that part of you, which is a person that is incredibly hardworking, that you are driven by generating employment, as you've mentioned, and by challenges rather than making money or the accolades of making it to Forbes under 30.

1:36:35So I very much hope that these young generations, young students that are looking up to people like yourself look and try to mirror and reflect that side of you, the human side, because it is the human Nicola that then created a great Nikola CEO and a great Nikola founder and not the other way around. Yeah, I think a lot of people... And yeah, the founder and CEO distinction is also quite different. I think that it's been over-glorified. I also think the economic environment of our times kind of like forces people to think about being entrepreneurs because life is getting harder and harder for like the average person in an average job, right?

1:37:16It's not the only thing. I think people should know why they're doing it you should find something you're unnaturally good at and double down on that and one big thing I believe in is doubling down on your strengths I would have loved being a diplomat when people ask if you weren't doing what you're doing and you could be something else not starting another version of your company what would it be? I would probably, this is the reason why I started this podcast, and I think I've never said this on camera. In my second life, I very much love psychology, and it's something that really always drew me in.

1:37:55I was fascinated by how the mind works, the combination of nature, nurture, what basically happens throughout our lives or our genes that end up developing a mind that behaves this way or that particular way. So definitely I would have been some sort of psychologist. The reason why I didn't do it is because I am quite ambitious at creating things and also utilizing my time very efficiently. And if I were to be a psychologist, most of my time would be deployed in helping this one single person or two people, which I don't think is the most useful way of using my most valuable resource, which is time.

1:38:34but if you if the question would be slightly different not if I if I had a different life but if I could have multiple lives yeah right this would benefit this would definitely be one of them yep yep yep yep what about you um yeah look I think that um yeah if you could save game in life and play levels with different things and stuff um I like to joke that if I was born filthy rich I would have been like a historian of like Yugoslavia and its breakup and like everything that contributed I'm still a bit of a like you know amateur historian I think that you know something like a career in like diplomacy and politics would have been really interesting right not for the sake of power and stuff but more for the sake of I think you can still do that within your life lifetime if you exit your company at some point soon well I mean I think that like not politics for the sake of like politics and power but more like the solving a historic threat and weaving it into direction.

1:39:33I think that's interesting. What are you good at? What do you know a lot about? You shouldn't know a lot about it. In my case, I love history. I love that side. Usually people have an interesting answer when you ask them. Like you, it's hard to marry what you do with that thing in a way that would double down on your strengths and then project it out to the world. But if you can find it, it's, I think, the ultimate expression of you. right and some people might not have that second thing they're probably the luckiest ones because they're like this is it i'm already there right thank you