Welcome to Spotlight On, a podcast about how companies are built from the people doing the building, one messy, exhilarating decision at a time. I'm Matt Wagan, partner at Excel, and this is Spotlight On. I'm excited to be joined today by Arkady, founder and CEO of Nebius. Welcome, Ark.
Thank you very much for having me. Here, I was thinking this week actually, coming up to this podcast, I was thinking about how you and I first met, and it was just over a year ago. The Excel founder Jim Schwarz introduced us, and I was also thinking about the fact that you and I have met each other across three or four different countries over the past year. And one of the things that's incredible about you is you seemingly live on a plane.
Where has that plane taken you recently?
Today it brought me here just several hours ago. Yeah, I live between different locations. My family is in Tel Aviv as well as several founders of Nebius in Tel Aviv. The majority of our engineers are in Amsterdam, so all the development and the back office is there. And here in the US, particularly in San Francisco and the Bay Area, is where the majority of our clients are. So I'm spread between at least those three places — Tel Aviv, Amsterdam, San Francisco — and all over, Europe, London, whatever.
Yeah. Well, I'm incredibly excited for this podcast and admittedly a bit nervous that we might not have enough time to cover what is arguably one of the most unique and fantastic founder journeys that we've encountered. So I'd love to start out with just hearing a bit of what you're up to today with Nebius. Tell us a bit about what Nebius is.
Nebius today is — well, it's formally a public company and NASDAQ-traded, but in essence it's a mid-stage startup. We're just developing our new business model. How we got here is there was a company established back in the '90s called Yandex, which eventually became the largest search engine of that region, of Russia. When we started, it was even before the internet actually existed. It was first computers and networks — local area networks, wireless networks and all that stuff. It was before Wi-Fi, actually; it was not called Wi-Fi. And then eventually we were building larger and larger networks, and then what we were building started to become an internet. And then one by one all those companies and vendors were actually purchased by Cisco, more or less. We became the largest Cisco distributor by this. And then inside this company, basically that was a value-added distributor of smart things where you need to train, you need to build a network of competent resellers whom you train. And then on top of it, we had several other initiatives, startups, business units, whatever you call it, where something else interesting was happening. And then when the internet actually came, we thought we'd demonstrate our product online, created a page, and then eventually this page became very popular and became a search engine. And then in 2000 we decided, well, it's kind of business, it's something big. This page became very popular, became number one, and that became an advertising business. And for the next 10 years it was advertising online services and then much, much more. Went public, a NASDAQ-traded company since 2011, again founded in 1997, a year before Google was founded. We were competing with AltaVista, then went public, became much, much more than just a search engine. All online services you can imagine — mail, maps, whatever — and then went offline. Became the largest taxi company like Uber, actually partnered with Uber later; there was a joint venture. And then went to all kinds of contemporary offline-online services — delivery, e-commerce and so on.
I think you're being modest actually, because in many ways it often gets called the Google of Russia, but in reality it was the Google, the Uber, the Instacart, the DoorDash, the Spotify, even the weather app as I recall. 25,000 employees. Quite an incredible journey, that story.
Yeah. Before the war, the Ukrainian war, it was $30 billion valuation on November 2021. And then the war started and half of top management and like a thousand engineers eventually left the country. And we started thinking, what else can we do? And we realized that actually several, we call business units or projects, which were targeted at global markets, now are more or less outside — like several startups. And also, formerly we were still a Dutch company and traded on NASDAQ. So after two years, we managed to technically sell the Russian asset. That was how it was implemented from the legal point of view. And so it was a Dutch company with four different startup ideas and a thousand engineers.
And so you're on this journey in Yandex and building this very large consumer internet business, but I think what most people don't realize about the Yandex journey is that although it was very B2C, Yandex Cloud was actually in many ways the early innings of your B2B experience. And so you started to think about maybe even before Nebius what would eventually be a lot of the core of what you would build with Nebius in some ways.
Yeah, Nebius yesterday when we just left, we tried to do something in the cloud area because there was no — the ChatGPT moment didn't come yet. It came just a year after. And when it happened, we realized that the assets we have, the people, the talent, is very well fit to this new era. And this is what we know best. Actually, our previous company was all based on AI, and the infrastructure which we built was hundreds of megawatts of data centers, and all the team who built it was with us. And we realized we can build this infrastructure for somebody else now.
So in a lot of ways you've almost been riding a continuous wave of — the internet comes, you build into search, you start to see the infrastructure to support that search, you carry that infrastructure forward now into the Nebius journey, and AI becomes the next driver.
Yeah. Part of the Yandex was very much focused and successful on one territory, but I always wanted — I thought it was underutilization of this resource. It was so many talented people working just for a great market, but it was just what, 1%, 2% of global market, right? It deserved, I thought, deserved more, to be exposed to a larger market. So in the last several years, since 2014, I was actually looking what else can we do in the world, how can we apply what we have in some something somewhere else, with which forms, which formats. And actually then when, unfortunately when the war happened, we cut it without thinking what we're going to be doing — just cut it, start everything from scratch. And then we realized actually most of the people who were working for this global project, they think the same and they are around us. And we realized that there was like a thousand of us, and that became a new company. And again, we didn't know when it started, we just left. We didn't know what it will be, what's next. And then it was 2022, the beginning of the year, and only a year later, end of '22, this ChatGPT happened and everybody realized that this will be the big thing. And then we realized, okay, we're relevant, we can do something here.
What was that journey like for you, bringing that team of both hardware and software engineers who I think rallied around you in wanting to, in this next chapter — what was that journey like bringing them into Amsterdam, bringing them into Tel Aviv, and getting going with what this next chapter would be?
This company actually — the main asset is the engineers and data centers and tens of thousands of GPUs of data centers which we built and continue building, as well as several other interesting startups which maybe we'll cover later. But this is the main business — the AI infrastructure. This is what we decided we're building now.
And so you find yourself in a situation where you're arguably maybe the most resourced startup we've ever come across. You are already a public company, relisting under Nebius. You have what I'll call four or five hundred software AI engineers inside the organization, and you have billions of dollars on the balance sheet ready to go tackle this, which I think many would argue is necessary to go chase the vision that you have. But talk to me a bit about how you came up with what the vision for the assets that you had inside of the Nebius Group — and obviously it's a hotly contested market, infrastructure — but why did you think that your team had a chance to win in that market?
So the business which we're building now actually stands on, I would say, three legs. One is people with experience and technologists. And we — yes, we have hundreds of engineers working together for many years, knowing what they're doing, with an experience of building one of the most advanced infrastructures in AI, which was actually one of the largest clients of Nvidia before all this AI and ChatGPT happened. So that's definitely an asset. Plus we had one of the data centers in Finland, specifically with us. So we had a data center, we had people who built infrastructure and Yandex Cloud in particular, and we knew how to do it. We had to recreate everything from scratch because we had no IP rights, but now we have it. We rebuilt the code and everything, and now we have team and technology which builds data centers very efficiently, builds racks inside those data centers which are also very efficient, and build the software stack on top of this racks, on top of this hardware. So it's a full-stack AI cloud. This is the first leg and we have it. The second leg was all this corporate and finance operation. We had to separate, pass through all the compliances, and eventually we became NASDAQ-traded. We didn't expect it to happen, but it used to be a NASDAQ-traded company, the trade was suspended for two and a half years, and then one day last October NASDAQ called us and said, 'Congratulations, you are compliant again, you start trading on Monday.'
