Stavros Papadopoulos: How Serendipity Sparked TileDB's Startup Journey (Part 2/4)

Intro - 00:00:01: Welcome to The Biotech Startups Podcast by Excedr. Join us as we speak with first-time founders, serial entrepreneurs, and experienced investors about the challenges and triumphs of running a biotech startup from pre-seed to IPO with your host, Jon Chee. In our last episode, we spoke with Stavros Papadopoulos about his early years, from his time in academia to the pivotal decisions that led him to explore new opportunities abroad. He shared how a chance encounter set him on the path to entrepreneurship and reflected on breaking into the us tech scene. If you missed it, be sure to listen to part one. In part two, Stavros takes us through the early days of TileDB, sharing his journey from research to industry and the challenges of securing funding and finding product market fit. We'll explore the technical hurdles of building a scalable database, the complexities of breaking into enterprise markets, and the strategic decision that shaped TileDB's growth. Stavros also reflects on the role of mentorship, the persistence needed to drive deep tech innovation, and how better data infrastructure can accelerate scientific discovery. 

Jon - 00:01:24: So when you made that decision to not embark on that five-year journey, where did you end up? Where did you go? 

Stavros - 00:01:32: Yeah, and it's interesting because with business, I did it. So it's a good question, right? Why didn't I do it with that, but I did it with business? Because with business, I went down an eight-year path now, an eight-year on something I did not know about. So we're going to talk about this later. So why? That decision, right? I made that decision on not going down the cryptographic world. But I did do the exact same thing that I avoided in business. I'll tell you about that in a bit. What did I do? So what I did was that I was trying to look for a job with my current expertise, which is data security, not security, data cryptography. You see what I mean? More applied towards data. And this is what I was trying to do, but completely serendipitously again, I contacted one of my collaborators. We had several successful papers together. His name is Minos Garofalakis. He's quite famous in Greece and beyond. He was in Bell Labs. He was a fantastic, fantastic career. And I contacted him. I said, you know what? I've been in Hong Kong for about eight years now. I really want to experience the U.S.. And if I don't do it now, I'm going to get married in Hong Kong. I'm going to stay home. It's either now or never kind of situation. Can you help me? Because he had connections and everything. He says, absolutely. He helped me a ton. But here's the funny thing. We had one of these big conferences of ours called SIGMOD. It's the top conference for databases. It's very difficult to get papers there. So I had the paper there. And I contacted him, I don't know, in May, but that was happening in June. So he says, hey, how about we schedule some meetings for you in June? I'm going to be in New York. It was in New York City. So let's meet in New York City in June. I'm going to connect you with my friends and something is going to come up. I'm going to tell them that you're looking to start, like, if not in the next semester, the semester after, because I had another year, at least for the professor position. So that's exactly what happened. But I had scheduled a massive road trip at the time with another two friends of mine. One of them was the guy who welcomed me to Hong Kong. So we took a car from New York City after SIGMOD and we made a huge circle in the U.S., a full circle. 22 days on the road. So I was in a vacation mode, so to speak. So I went to the conference. I did make the connections. I took meetings and stuff. And then I was on the road. And on the road, sometimes I had the internet. Sometimes I did not have internet. You know how it is. Sometimes we're driving for tens of hours at the time. Sometimes we're lost somewhere. So I received an email, actually two emails from Minos, introducing me to two people. One was a big boss at AT&T Labs that he had connections with. And I had some friends there also. And one at MIT. Sam Madden from MIT. And they both offered me an interview on the spot. Paragon in Santa Fe. And he said, Dude, like, when are you going to be here? One in New Jersey, one in Boston. So I took out my map. We have maps. No, there are Google Maps as well, but still we have notes, right? Or I mean, how do we make it work such that on the way back, we stay a little longer in Boston and one day in New Jersey that we had not sculled because we had to catch a flight.

Jon - 00:05:09: Yeah. 

Stavros - 00:05:09: And we make it, but in Santa Fe, anything can go wrong. Yeah. It's very far away. We were. I think 16 days away, something like that, 16 days away. So we say, okay, that date in Boston, the next day, New Jersey. So, and I tell everyone, you know what? Like, especially the North, we're going to drive for 32 hours straight. Otherwise it can't happen. And we made it. So I gave an interview at MIT. There was some other Mike Stonebraker. These are giants, okay? Like giants in databases. And effectively, the MIT position was a dual appointment, so to speak, because... Intel Labs was looking to hire, but that person was embedded into MIT working with MIT persons. Like the office is there, then, you know, their research is done with them. You're effectively at MIT, but getting paid by Intel, which is great. Getting paid by industry, working in a top university. 

Jon - 00:06:05: That's rare. I feel like that's very rare. 

Stavros - 00:06:08: Beautiful. 

Jon - 00:06:09: Yeah. 

Stavros - 00:06:09: Intel is amazing. So anyway, I give the interviews. They go surprisingly well, especially at MIT, because at MIT, I thought I knew the people I was talking to, the giants. And I was like so happy to speak with them. I was very relaxed because I said, okay, I'm going to go to AT&T. I'm really hoping I can get the AT&T one because I had friends there. They knew my work. And also the work I would do at the time at AT&T was extremely relevant to what I was doing. That's exactly what they wanted to do on privacy, security of data. MIT wanted to do something completely different, like systems. And I was a good programmer. But systems is a little bit different. And I mean, people listening probably understand. Like, you can be a good software engineer, but then there are other software engineers who are professionals. And it's different. Like when you're developing systems, I mean, still it goes research, so it goes okay. But if you're developing systems for production, and I know from the people I hire, right? So I know that they have to be way better than me. And they are. They're different. Like the same thing that I was talking about, the cryptographers, it's the exact same thing for the software engineers. They have spent an eternity being extremely good at that, building systems, not for research, but for production. And there is work there to be done that you might not like, right? I was building code for algorithms. That I could do extremely well. Like find the best algorithm with asymptotic complexities, making it work very well on the hardware. Yeah, but again, that's not software engineering. That alone is not software engineering. It's more. It's about deployments. It's about the reliability. There is so much that goes into software engineering. That's why I respected that much. But at MIT, I knew who I was talking to and I was telling them, listen, I want to work on these problems and we'll discuss about this. I believe I can do very good work there, a very good job. And I was very candid. I said, I don't even know this area. If you see my CV, I'm good at algorithms and I'm good in general at, you know, learning very fast, and applying this knowledge. You know, if I was here, I would do an extremely good job because I would be very inspired by the environment. I'm extremely hard worker. I'm going to learn very fast and I'm going to show you some success. And they gave me an offer. Like I also spoke to the Intel folks, of course, and I told them the same. And they got intrigued and said, you know what, like if you're willing to learn. And this is so unique in freaking U.S., man. And that's why I've been in so many other places, but it is unique. Yeah, they will give you a chance. They will give you a chance. It's up to you what you're going to do with it. But it's not impossible. Like a kid from Xanthi, 60,000 population, like born into a family that has nothing to do with this. You know, the universities and all that stuff. Ending up at MIT at Intel. And also for me, it was extremely important because remember what motivated me was that magazine. The magazine at Intel there, at MIT. So impossible to pass on this. Impossible. And I was good in Hong Kong. In Hong Kong, again, I had the idea of going to the U.S., but if I couldn't, fine. Hong Kong, I was having a ton of fun. It was good. But for that, oh my goodness. I told my wife, I said, there's no way I'm missing out on this. Even if I fail, I can't care less. I'm going to go for a year. I'm going to come back. But I'm going. I'm going. And it was instant. Like I took the job. I said, fine, I'm here. 

Jon - 00:09:35: I love that. And I think I was telling Christina earlier that I've only been in the Bay Area my whole life. I probably take for granted what you were just describing of like. Being given a chance. When I probably didn't deserve it. 

Stavros - 00:09:52: I mean, I thought I deserved it, but the thing is, just the way I think about it, I did work. And you're going to see later that after the fact, I was in the right place, even for them and for me. I think it was a good match. But the thing is the probability. Like, I'm thinking about it statistically. Even if you deserve it, the probability that you're going to give a chance like that is very small because of the competition and because of, again, serendipitous events. Like, a lot of factors play into this. And frankly, somebody who deserves to be there will be there. I promise you that, man. If somebody really deserves to be there, it might take longer. And by the way, by deserving, I mean not the talent. The talent alone. doesn't suffice strategy. Like, they have the talent, but also they're targeted. Then nobody's going to come from MIT and say, please come work for us. There's no way.  

Jon - 00:10:48: Yeah, yeah. 

Stavros - 00:10:48: Or any other top universe. No way. You need to be strategic. You need to really want it. And at the same time, the opportunity needs to arise. But if you keep on doing it, the opportunity will arise if you deserve it. So that's how I'm thinking about it. Definitely luck plays a big role, but you need to do a couple of things to get this luck. I called Minos. 

Jon - 00:11:10: Yeah.

Stavros - 00:11:11: I called him. I said, Minos, I'm thinking about this. Minos had the connections. Then I did very well in the interview. I did well. You see, like, you need the luck, but you need to chase it, man. And you need to perform. You need to deliver your end of the deal.  

Jon - 00:11:30: It's kind of that saying. It's just like, it's just being prepared. Like, you just never know when that opportunity will arise. 

Stavros - 00:11:36: I think you need to be proactive. Not just prepared. Proactive. So yes on that. Yes, I agree. But don't wait for the opportunity to arise. 

Jon - 00:11:43: Yeah, don't wait. You knocked on the door of your professor back in Greece. 

Stavros - 00:11:47: Oh, I definitely made it happen. And by the way, this is how business happens as well, a hundred percent. It's the same thing. It's not an accident. And also the people I hire, the people I hire, I see trades like this. And the people I hire, I see them. I see them being proactive like this. And are there more talented people around? Perhaps. But talent alone is not useful to the company or myself whatsoever. You need to be proactive. You need to be aggressive. You need to be hustling. Otherwise, you can't do, especially startups, more mature organizations perhaps. And even there, I think if you want to succeed, you have to be proactive. But you need to be a hustler, man. If you are not a hustler, things are not going to happen. And the talent is on the sustainability side because somebody may be a hustler and you may be given opportunities and take them, but then waste them because you're not good. Because you, it's not good. It's vaporware. 

Jon - 00:12:40: Yeah. 

Stavros - 00:12:40: Like you present yourself great, but there's no substance. So the substance is important for sustainability so that you, and the consistency, that you are consistently good. Not just something that was like a comet, like a passing star. 

Jon - 00:12:56: Yeah.

Stavros - 00:12:57: And then this evaporated. No, you're consistent. You need to be skilled and consistent.  

Jon - 00:13:04: Absolutely. And I think likewise, when, when I'm doing interviews, it's like, I always think about hiring for attitude and then training for aptitude. Like it's hard to like. Train hunger, like kind of what you're, what you're talking about, you kind of just like have it or you don't, but it goes back to what your professor at Hong Kong said. It's just like. If you like it and you're capable, like typically that hunger is like, if you like something, you're going to like go, it's like a, you're compelled to do that thing. You're compelled to knock on that door or call up your friends or your network. And, and I think also too, just like picking up the phone and just like calling is like such an underrated like skillset, like exactly what you said. You're like, Hey, like I'm interested in this. Do you know anyone that can help me with this? 

Stavros - 00:13:52: Oh, yeah. But you need to be targeted, right? There's a famous video going around with Steve Jobs from 30 years ago saying exactly that. But that came after I did everything. Like I saw it, I don't know, last year, right? I said, yes, man, this is so true. I hope people that are watching it act upon it. It is true. It's exactly what we're seeing here. Like Steve Jobs was saying exactly the same. He has a very beautiful story about Hewlett Packard, what he did there. I could see myself doing that for sure. I've done it. I didn't go to the equivalent of here, but yes, I'm doing that 100%. So it's definitely underrated, 100% underrated. But what I advise also, because I have seen, I'm being contacted by so many people. I have a big network and I'm trying to be active as a mentor myself as well. What I do not like is open-ended questions. Somebody is asking, hey, how do you feel about the AI domain and what should I do? I cannot help you. I can't. Like, first of all, I'll gossip with my friends. I have opinions. But that's gossiping pretty much, if you really think about it. Like, it's all philosophizing about the AI, what's going to happen. Yeah, that's a fun conversation to hear, but that's not actionable. I need to be helpful in an actionable capacity. Do you want a particular warm introduction? And I need to believe that because I respect a lot those introductions, my friends did them for me. So I have to be very careful. My friends trusted me because they said, I trust Stavros. He's going to deliver. He's not going to expose us, right? I want to feel the same. I want to feel the same. So if I am to do something like this, I really need to believe in you 100%. And don't take it personally if I don't know you. I don't know you. If I know you and I believe in you, I'll do it. I'm always very accessible to help because I go. I received help, but it has to be actionable. Like something specific. Like, can you please look at this deck and let me know if my assumptions on go-to-market strategy are valid? I've done this numerous times. And they thank me. They call me. They say, hey, I changed this. I got funding.

Jon - 00:15:58: Yeah, yeah, there we go. 

Stavros - 00:16:00: Okay, I'm happy. Great. I've done the same. I don't need anything in return. I mean, they're saying, hey, can I get you some advice? No, it's okay. Like, call me. I'll give you, like, give me a specific question, I'm going to answer. I'm going to answer that question.

Jon - 00:16:15: Actionably. I feel the same exact way. And it just reminds you of when you sent that email right before you went to grad school and you got that instant response. Same thing. And nowadays, sometimes you got to get through a lot of messages. But I think people don't realize that if you come in with an actionable question, you'll be surprised at who responds to you. And I think there's a special thing about this. I don't know if it just is purely an entrepreneurial thing, but it's just this kind of pay it forward, give it back kind of thing. I always remember exactly the same thing what you're describing, who's mentored me and who's kind of responded to me and maybe made that connection that I would not have had without some help and how much that kind of really just was a massive inflection point in my journey. So it's always this thing where I feel not just compelled, but it's almost like a civic duty where I'm like, you got it pay it forward. 

Stavros - 00:17:12: It really is. It's the only scalable model. Think about it. It's like that's the definition of scalability, the definition of scalability. So it is a civic duty, to be honest, especially if you and even if you have not been helped, do it. Be the root of the tree of the tree, right? That expands. Be the root. You don't need you don't need to wait for somebody to help you in order to go help somebody else, especially if it doesn't cost you too much. Like in terms of time, I can spend half an hour looking at the deck or 15 minutes, whatever it takes me. I'll do it. I don't mind, as long as it is kind of actionable, especially even for decks. I'm seeing so many decks, like as many as I've received by now. But, because I'm not a VC. I need to be able to help. It needs to be in my domain. I need specific, like tell me exactly what to look at. I'm not going to fix your entire pitch, but give me, like, as I said, go to my stuff that I have seen and I have learned the hard way. So. I can bring value there. And I don't like posting on LinkedIn unsolicited advice. Like, oh, I'm seeing that all the time. It's like crazy. That actually intimidates people. Like somebody who is trying to break into the market, they have a little startup and they're seeing somebody be obnoxiously saying, oh, I did this and I got to $100 million ARR. It's not helping them. That's not what they're looking at. They're trying to hear from you how you succeeded and the lesson. Like it has to be phrased completely differently. If it is unsolicited advice, to aspiring entrepreneurs, you can't reach them that way. It's different, like, tell them about the hurdles that you had. Tell them about a specific example. Like your advice cannot be generically applied to every startup. It cannot. Now we're getting a little bit into the business and proactively, but it can't. You can explain what you did for your specific industry and with your specific funding situation, that too. One startup from another, difference in multiple respects and funding is one of them. If one doesn't have funding and the other doesn't, you have a lot of funding. Yeah. With a lot of funding, you can do strategies that are fancy and funky and they worked out. But when you are resource constrained, you don't even fathom doing something like this, because if it doesn't work out, it's death. So anyway, all I'm saying is. The help needs to be actionable and it shouldn't intimidate it should help, is supposed to help. 

Jon - 00:19:29: Yeah, it's like it's also very like what you're describing is like is like contextual, like it has to be within context. 

Stavros - 00:19:34: It has to be. Again, that's why I'm not posting generically. It has to be absolutely contextual. My startup is so different from yours. And there are so many ways that each of them can succeed or fail, but so many different ways. There are some generic principles that everybody should learn, and we can talk about this, what I did in a bit, when we get to my entrepreneurship journey. The generic knowledge, not advice, generic knowledge helped. But then the rest of the details, you need to see it in practice. There's no other way. You need to see it in practice. 

Jon - 00:20:08: I completely agree. And I think there's so many permutations and there's so many factors and levers that are being pulled. So I exactly agree. Like when, when folks are just like profitizing, like, this is the one secret that's going to like solve all the ills and you'll get the a hundred million ARR. You're like, what? Like, this is a, how, like how, like I'm a leasing business and you're, you're in software. These are like two very different businesses. But okay. I'm getting, I'm getting distracted. You're at MIT and you're now, and you're, you're getting paid by Intel. Sounds like a sweet gig. Can you talk a little bit about the work that you did? How was that experience? And then you're kind of alluding to when did you end up deciding to jump into business and take that eight year kind of detour to level up to par? 

Stavros - 00:20:59: I will explain and you will see that there was another serendipitous event. As serendipitous as it gets, as crazy random as it gets. Not that the event was random, but the outcome was not random. It would happen. It would happen. So from the start. So first of all, when I got into Intel and MIT, the people I met were amazing. Like the smartest people on the earth. It's not random. The fact that MIT is MIT and it's not random that Intel is Intel or has been, throughout the years. The organization they are. There are a lot of smart people, gathered in these small places. Now, how they got, they gathered there, it's a matter of their founders and, you know, maybe other lucky incidents. But the fact is that they gathered, they're very smart and they're very hardworking, very hungry, very talented. And that's why you see what you see from those organizations. So I was very lucky to be there. And I started working again on database systems. I wanted to build something new, but at the time, trying to see what other projects were around. So I tried to work on some other projects. Eventually I ended up proposing to work on TileDB. Like that was my first project. But that was because I was identifying some very technical, very low level, very uninteresting to the broader audience problems, as we always did with research. Okay, we found a very narrow, very narrow problem that needed solving. And I set out to solve it. And I'm not going to get too technical unless you want to get into the details. But the whole point was. But I was not very happy. With the majority of database systems out there that are tabular, like they store tables called relational databases. And the reason I wasn't happy from a research perspective is that I always thought that we did a lot of work there already for the past 50 years, 50 years worth of work. So I said, I can't differentiate myself that much there. I wasn't thinking as a businessman at the time, but what kind of contribution am I going to make there? It's going to be even narrower than what I thought. And the competition is crazy. Everybody's building tabular database systems for a good reason. The market is big and it solves a very important problem. So anyway, I was interested in problems that are not tabular. And I proved that also during my PhD that I was not working on tabular database, I was working on geospatial, like spatiotemporal databases and the security stuff. So I always wanted to think out of the box and do something else. Now, at the time, Intel was very heavily invested in machine learning, not the AI stuff that you see here, like pure machine learning, right? Optimizing their software on that hardware that they were building and vice versa. And at the same time, at MIT, people had the extreme knowledge of databases. Extreme, though. They had built many companies, sold many companies. So actual databases, right? Not just research. Research and commercial products. And I wanted to combine the two. So effectively, I wanted to build, at the time, right, before AI was crazy, kind of an open database for machine learning. But then I took a very bottom-up approach. Other people took a top-down approach. More recently, I took a bottom-up approach. I said, we need to start from storing the data. And machine learning works on matrices. It's not working in vectors. It doesn't work on tables. So why are we storing data in tables? Let's start storing data in matrices and vectors. So I looked around, there have been a couple of other systems doing this, but without getting into a lot of details, they did not meet my requirements for a bottom-up approach to building a database system that is capable of doing machine learning, advanced analytics that popular databases cannot do, and capture use cases beyond customer transactions and stuff like that. Like, I wanted to do DNA sequencing, like to handle DNA sequencing. I wanted to handle all in genomics and life sciences. I wanted to do the geospatial stuff, and the popular database could not handle that stuff. It couldn't. And I saw the cloud rising, so I saw a lot of potential of storing data on object stores like S3. So I was exploring very different architectures. I could not leverage an existing database. What I was trying to do was radically different. Now, almost everybody is doing it, but at the time... No one. Like the separation of storage and compute, you know, we're storing files on S3, but then the compute happens independently. I was building that before it became modern and now standard. That's the other problem, unfortunately, with technology. There are a lot of smart people around. They're working in parallel with you. They may bring stuff to market faster, right? And this is what happened in my case. But this is exactly the architecture I had built, at least internally. But that was it. That was it. The good news for us is that I was working at Intel, and I presented my ideas very radical. All of those ideas were radical. At the time, using multi-dimensional arrays for capturing this, using a heavy use of the cloud, right? The separation of storage and compute. And I presented to Intel, the big Intel event, and there were some people from the Intel Health and Life Sciences division, a small division working in Health and Life Sciences, who were working with a Broad Institute. A lot with the Broad Institute. And they told me, you know what, like this structure might be able to solve a very important problem that the Broad is having and everybody's having. They have a lot of VCF files, a lot of variant data. And the tools they have today, they're not database tools. They're reaching scale limits. Can you look into the problem? So the Broad invited me. It was just across the street from my office. So I went there. I discussed with the newly formed data team there. And they were telling me we have 60,000 files coming in. Up until then, they were dealing with 3,000. Now 60,000. And we want to store them on the cloud, but there's no cloud native solution. And we have all these issues. Can you help us? And that's exactly what I did. The very first use case that I applied TileDB on VCF data. We had a big success there. We solved a very big scalability issue there with Intel and abroad. The code was a first step. So it was amazing. It was called GenomicsDB at the time. That's how a lot of bioinformaticians know it as. It's part of GDK and all of that stuff. But it was a first step. And then I was thinking about how to refactor it all and make it more native. But at least there was a proof of concept that this is going to work. It solves a big problem. And I was kind of proving that, first, there is a kind of a market for databases in life sciences. But up until then, nobody was exploring. And second, that we can handle data with a database that no one else could handle before with a tabular relational model. So there is a need. And more sophisticated because it's more difficult than the relational model because it subsumes it. It's a superset. So it includes everything in the tabular and more. And with that, first of all, I proved a lot of people at MIT and Intel to pursue it. We wrote a VLDB paper, like in one of those big conferences. So we proved the technology. My code was not system-level code, but Intel helped me. So again, as a proof of concept, it was a proof. It wasn't just an idea on a deck. It was like, yes, this is working. There's a new market. The market is growing. It might be small, but it's growing. And it is worth investing. And here's where the serendipitous event comes up, so I have always been thinking about, how would I make this as a company? What not, but I knew not, not, I knew, zero about business. When I say zero, I mean, and thankfully I knew what I did not know, very well, so it's not that I said, yeah, it's easy, I said no, like, this is a new territory, that's not just, my PhD is not gonna help me here. 

Jon - 00:29:08: Yeah.

Stavros - 00:29:08: Like, I knew it. Whereas others are saying, but, I mean, PhD I'm smart, it doesn't matter. It's a different domain, you need to learn, you need to learn, and you need help. So anyway, I was thinking about it, but I was not acting upon it because I knew nothing about how to create companies. Zero. So what happens is that. MIT in Stata, there's a canteen where everybody goes for food. So I was going regularly there when I was at the office, I was going to the canteen to eat my lunch. And in the elevator, I found another friend of mine, Greek, working in another research group. He says, hey, what are you doing? I said, I'm going to the canteen for lunch. He says, me too. And I'm meeting with another friend. He told me his name. I said, oh, I know of him, but I don't know him, another Greek guy. And also a researcher in another group at MIT. He invited me. He said, okay, come and eat lunch with us. You're going to meet the guy. So I'm meeting with a guy. His name is Stelios Sidiroglou-Douskos. I asked him, what do you do? He says, I'm a researcher, but I had a company. I exited the company. Now I'm doing a little bit of angel investing. And then I explained to him what I do. And he says, dude, this is a textbook startup. Do you know anything about startups? I said, I know nothing. He says, come by my office later today. So I go to his office and he gives me a couple of books. He says, read these books, come back. One of them was Disciplined Entrepreneurship. It is a good book for someone who knows nothing. Like if you don't know anything, it's a good book. And the other one is, I think Zero to IPO from Frederic Kerrest, who is also an investor. He's the co-founder of Okta. So these are the two books that I quickly recommend for somebody who knows nothing, nothing about entrepreneurship. So anyway, he says, listen, man, can you read those books and come back to me with some kind of a deck? Because these books kind of advise you how to, it's not a primer for pitch decks, but there's also information on the internet. So he said, go research, like to see whether I would do it, right? And I do come back with a deck. Of course, the deck was not very good at the time, obviously. But the bottom line is that I told him about this. He says, dude, you know something? Do a company. Like quit the company. I'm going to invest. And it was a good amount of money. It wasn't 10K. It was a good amount of money. The guy did not know me, by the way. That was like the third time. You just made a lunch. Just made a lunch. Right? It was like the second, third time. He said, I'm going to help you. I'll help you. And I'm going to introduce you to my friends in Silicon Valley. You need those. And through those, so we're going to get a couple of other angel investors. And through those, we're going to go to VCs. And then he said, it's going to be good if you get Intel to invest because Intel has Intel Capital. But Intel Capital doesn't invest that early. A big fund, like later on. But my managers were extremely supportive. And they said, you know what? Like Intel Capital has, because, I explored doing it in-house, Intel. And the issue was that Intel is a hardware company. There are, unfortunately frequent reorgs, things get shut. If it's not a cash cow business. 

Jon - 00:32:13: Yeah. Non-core. Non-core.

Stavros - 00:32:15: Non-core? 

Jon - 00:32:16: Yeah. 

Stavros - 00:32:16: A new manager may kill it. And I didn't like that because I was planning on working very hard on this. So imagine if I put in all this work for a random reason, unrelated to your product, when the business gets shut down. So I didn't want to take my chances. But they said, if you do it, we're going to support you. So they introduced me to Intel Capital. Intel Capital had a very small initiative for university spinoffs, very small amount. But I wanted the validation. I wanted to make it an Intel MIT story, which it was. And it is. But Intel decided to invest. I raised a $1 million round. Trust me, this is nothing for databases. Now we're near $1 million. I'm surprised we survived that long. And then continue to raise, of course, because $1 million, it's a non-starter. Today, I wouldn't advise anybody to start a database business with $1 million. Forget about it. $50 million and above. And if you can't, don't start it. If you can't raise from the get-go, if you don't have connections and stuff, $50 million? Do not start a database business. It takes a long time, very expensive talent. It's a very hard job to do. But, and I'm saying that because I still had a plan. I said, I will not be able to raise that much money. I don't have track record. I can't convince somebody to write a $50 million check. There's no way. I don't have the data points also. Let's be realistic. I have the Broad, but the Broad is just a user. It's not even a customer. And it wasn't. I have validation, technological validation, but I don't have anything else. So a million dollars, it is. When you have only a little bit. A little bit of technological validation, nothing else. So it was fair. In other words, like I couldn't have raised more. But I had a plan. I said, okay, with 1 million, I can prove that I can transition this. Little project that I got into something a little bit more serious. Like I'm going to find some more professional software engineers better than me to polish it. And I'm going to find one customer. And that's exactly what I did. So I spun it out. I started, I read like crazy, right? Like those books were nothing. Like, I mean, I read another 200 books within six months to prepare and be ready. How do we incorporate? How do we close deals? How do we write the contracts? I have a list for this kind of books that you must know. And this is like, it's a no starter. If you don't know this, don't do anything. And then if you read them, they don't guarantee your success, but it's a sufficient condition, but it's necessary. Necessary, not sufficient. So that's what I did. I found some very talented people. Unfortunately, I didn't have a co-founder at the time. And this is a mistake. Try to find co-founders. For me, it was like, we're going to figure everything out on the fly. It's not ideal. But we did. Like I did put in the work, the work so that we do figure it out along the way. And we scale gracefully. We start raising more and more and more money in order to build it, mostly to build the product. Like it's, it's expensive to what we're doing is, is very expensive. And I will, I will talk about what eventually the vision shaped to be. But it was that serendipitous event that gave rise to this. Also, I had a friend in Greece who said, you know what? I'm going to support you too. So I said, okay, with these two friends, I can quit. I'm going to get a big pay cut. 

Jon - 00:35:30: Yeah,

Stavros - 00:35:31: For sure. 

Jon - 00:35:31: I was going to say. Yeah. 

Stavros - 00:35:33: And at the time I, it was getting worse because my wife was laid off at the time. We had bought a house. We had very high mortgage for our standards. And she got pregnant. So all of that while I was incorporating and raising money. 

Jon - 00:35:48: Oh, my God.

Stavros - 00:35:49: And I was telling my wife, Mike, I don't know how we're going to survive. I don't know. 

Jon - 00:35:54: Oh, my God.

Stavros - 00:35:56: I mean, what else? I don't know what else would be worse, as long as health is not related. You know, like other than health, I don't know what could be worse than that. 

Jon - 00:36:06: Yeah, yeah. 

Stavros - 00:36:07: Like we were barely making it month after month after month. But this is pretty much how it started, Jon. Like this R&D business event and some other angel investors that got in, other VCs that got in, like Nexus Venture Partners, who I really love because they're helping us throughout this funding process up until today and they're at the board. But they came a little later. They were the last to come in, but they are the most trusted, among the most trusted funds that took us all the way afterwards. And that's how you raise money. Angels who can vouch for you, two VCs, VCs who are going to do due diligence, they're going to take a flyer on you, especially if it's a small amount, they will take a flyer on you. And if you keep on delivering milestones, they will keep on funding you or helping you to get funds from elsewhere. That's exactly what happened with us. 

Jon - 00:36:55: That's super fascinating. And also I just love how it was just like a, it was a lunch that just kind of just immediately sparked it. And I guess just really just like quick before kind of talking about the TileDB kind of like. From the ground up what you've been building. But it sounds like Intel was supportive of this kind of like spin out. But it was also, it sounds like you were also at MIT. Did you have to get the blessing of the MIT like tech transfer to do that? Or were they just like, eh? 

Stavros - 00:37:26: By the way, that was calculated, right? Because otherwise there would be hurdles and I would not be funded. Though, that was part of due diligence, like what can we do? The good news is that because of the nature of the collaboration that I was under between Intel and MIT, there were very specific rules. It was a formal, it was an official collaboration. It was like a center of Intel at MIT. So it came with bylaws and a lot of stuff. You see what I mean? It was official. It was not just Intel sending me to MIT. No, that was an Intel-MIT position with specific rules with respect to what IP I'm touching. So it was stipulated that everything that I personally touch is under MIT License, open source under MIT License. As permissive as it gets, I could do whatever the heck I wanted. 

Jon - 00:38:12: Wow, that is incredible. That's incredible. 

Stavros - 00:38:15: Otherwise, I wouldn't have considered it at all. So MIT didn't care. It's good for MIT, especially with some successes that we've had in the past and hopefully in the future. MIT can say, look, TileDB got spun out of MIT. Yet another one of their numerous, numerous successes. So, you know, keeping the tradition. For Intel also, look, guys, we have this mentality and all of that stuff. So no strings attached. Also, I kept, of course, the relationship. Sam Madden, very famous professor there, sits on my board. He's a very good friend of mine. And we wrote the first paper together. So I really wanted to involve him as well, like because I consider him as part of this. So he's busy and very successful. You understand, right? I wanted to do it because we wrote the initial paper together, a lot of ideas together. I get a lot of mentorship from him. And he's been with me throughout all this entire journey. We're talking about eight years for the company. 23 years before that, 11 years now. So I love this kind of relationships and I want them to materialize in a very meaningful way as well. So in other words, we structured it in a way that everybody was really happy and TileDB's success is everyone's success. Not that they need it. These people are very established, but it doesn't matter. It's yet another success in the tradition of successes of those two organizations. 

Jon - 00:39:40: That this is kind of like you designed like a kind of like a win-win win for everybody 

Outro - 00:39:47: Thanks for joining us on this episode of The Biotech Startups Podcast with Stavros Papadopoulos. Be sure to tune in for part three, where we dive into the evolution of TileDB, how the team scaled the technology, expanded its applications, and navigated the realities of enterprise adoption. Stavros also shares insights on building a strong engineering culture, the challenges of educating the market on new technology, and the strategic choices that have driven TileDB's success. If you enjoyed this episode, subscribe, leave a review, and share it with your friends. See you next time. The Biotech Startups Podcast is produced by Excedr. Don't want to miss an episode? Search for the Biotech Startups Podcast wherever you get your podcasts and click subscribe. Excedr provides research labs with equipment leases on founder-friendly terms to support paths to exceptional outcomes. To learn more, visit our website, www.excedr.com. On behalf of the team here at Excedr, thanks for listening. The Biotech Startups Podcast provides general insights into the life science sector through the experiences of its guests. The use of information on this podcast or materials linked from the podcast is at the user's own risk. The views expressed by the participants are their own and are not the views of Excedr or sponsors. No reference to any product, service or company in the podcast is an endorsement by Excedr or its guests.