EP.017: Time to take profits in NBIS or is it the next AWS?

[MUSIC] >> Sonan, Nebius is a $33 billion company. It's EPSN free cash flow negative. It's up four to five X this year, mostly as backlog has jumped. It's now your largest position at around 15%. Are you sticking or taking profits? >> I'm sticking. >> And I walk you through why? >> One of the core pillars of the deep research we do is field research. I'm using expert networks. I always start with AlphaSense. They have a great expert library, and they also have AI technology that allows me to get to the information a lot quicker. They allow me to access local periodicals and uncover information that I otherwise wouldn't have access to. They are used by 75% of the world's top hedge funds and 85% of the S&P 100. Not only do you have access to the library, but they'll also go find experts for you based on what you're looking for. It is one of my go-to ways to do primary research. Make sure to check it out with the link in our show notes. >> Hello and welcome, everyone. My name is Doug Garber, and this is Pitch the PM, where we debate high-conviction ideas from emerging managers. With us today is Sonan from Amber Road. Sonan, can you give us a quick update on your background and how you started Amber Road? >> Absolutely, and please to be here, Doug. We've talked stocks in the past, and I'm glad to have an opportunity to share my background process and ideas with you again and with your audience. My background is, I'd say relatively unique. I got my start in the world of finance in investment banking. I was one of the last survivors at Lehman Brothers that was able to move over to the investing side in private equity, landing in Asia, trying to convince founders and shareholders to sell us their businesses at the bottom of the market. And I'll say that having spent time in the smoke fill rooms of deal-making in the emerging markets has given me an interesting perspective on public market investing, which I managed to get to a few years after spending a few years in private equity in Asia. I've worked at Tiger Cub type of hedge funds. It was a technology investor at one of the largest emerging markets focused hedge funds based in New York called Emerging Sovereign Group. They've since closed a few years ago, but I caught the Chinese internet bubble 15 years ago. And then I was able to catch the US technology bubble and I shouldn't say bubble, that's a sensitive word these days, but the transition from on-premise to cloud computing, the transition from desktop to mobile internet, from legacy media to social media. And then today, having spent the last 17 years in the investing world, having launched Amber Road, which is a long-short equity fund based in New York that's focused on technology companies internationally. And that's companies based everywhere outside the US that serve global markets. And they're in the business of AI, of software, Fintech, internet, and other technology sub-sectors. And then the goal is to find the next generation of technology winners, which aren't going to be from the usual places. And they're not going to be from Silicon Valley potentially, they're potentially from other places. And so we're sitting here in New York looking globally. So emerging markets is an asset class where the starting point that many people have is thinking as a macro investor, or thinking thematically. They're the idea of, there are billions of people out there that have lower incomes, but they're growing. There's hundreds of millions of young people in XYZ geography, and then people tend to hear about the successful bets that famous macro investors make about presidential elections in Argentina or China reform and opening up. My starting point is different, and the reason why it's different is formed by my experience at stockpicking focused hedge funds, as well as, I would say, absolute return funds like Citadel. We are a stockpicking strategy focused on dispersion across between our long and short portfolios. And it's still agnostic. So I've made money as a growth investor, as a garden investor, as a value investor, on both the long and short side. And being able to avoid significant factor tills is something that, as you remember, is an important part of running a business at a Citadel, or another kind of leveraged hedge fund model. But I've taken that portfolio construction, risk management, philosophy, and kind of glued it to something to my longer term private equity style investment process. My first job, as I mentioned, in investing was in private equity, building relationships, discerning whether or not businesses were high or low quality. It was a non-calling quarters on earnings or trading based on short-term data. It was based on being in the room with management teams and figuring out what was going on on the ground in these various international geographies. And taking that stockpicking process and gluing it to a factor-aware style agnostic portfolio and risk management process that Unite engaged in, when we were at Surveyor, enables us to have this return stream in this portfolio that we will own long-term winners, but we will be able to stay put in these names through the volatility of the international equity markets. So it sounds like at the core of everything you do, it's deep research and really understand in the business model and the companies and management teams behind them. That's right. And I'm a technology investor, first and foremost. Where my first job at Lehman Brothers was in the technology mergers and acquisitions groups. So I've been doing technology investing before it was fashionable. And kind of in the wreckage of the.com crash, kind of the mid-2000s, technology was not a popular place to be and that's where I wanted to be. And so it gave me the foundation to look at tech companies in many countries in the world through their successful periods of time and their unsuccessful periods of time. And it's informed my ability to underwrite using case studies, using a research process, and be able to learn new technologies. Today it's AI before it was cloud, before it was mobile internet. And it made me able to underwrite that and invest in this sector, which I think is incredibly dynamic and exciting. So I really wouldn't look at any other sector. Yeah, it's a good sector right now. And it just shows that domain knowledge is cumulative. And that alone could be your advantage or your edge is just having seen all those cycles and experienced it. That'll give you the conviction to know what's happening this time. In international markets, technology is also early. If you think about e-commerce penetration, software, digital transformation at small or enterprise businesses, it's certainly AI. It's early for everyone globally. So there's always the next frontier in technology, whether that's geographic or product and technology cycle driven. There's always something new to learn, to build a conviction and then to navigate as they become public companies and go through the ups and downs of the markets. Before we get into the stock, we're going to talk about today, Nevis. I want to just talk about your idea generation and what your process looks like. What are the questions you ask yourself? So the fundamental process is very much similar to what we did at a Citadel or what I previously did at the Tiger Cup type of complex, that style of nesting. I start with a sector coverage model. So we build relationships on an ongoing basis with all the public companies that we follow in our space. We maintain detailed financial models. We build relationships across the channel for these companies. If it's a software company, we'll build relationships and get to know the distributors, the customers, the competitors, whether they're private or public. I spend a lot of time with venture capitalists in various geographies in the world, in Asia, in Latin America, in Central Asia, and learning about what they have their portfolios and how they're going to come to the public markets, but really how they navigate competing against public illicit companies. So over the last five years, there have been more new publicly illicit companies and technology outside the US. So there ever have been 15 years ago, if I wanted to start an international technology fund and really just be a Chinese internet fund. But now you have software, Fintech, AI companies in Brazil, in Southeast Asia, India, Turkey, Kazakhstan. So it's a very diverse universe. And requires systematic coverage. So that's the building blocks. It's building this base of knowledge and history with the public companies, the liquid public companies in our universe. The second piece is our-- so you asked the question about the question you asked is, I sit in New York. I have limited resources and limited time. So where do I spend my time? That's an important question for idea generation. How do you spend your time? And the question I asked is, too full. Why is there an opportunity here? And why am I best positioned to capture the opportunity? We've both worked at very well-resourced firms. We've seen how people with unlimited research budgets and can talent budgets can conduct public markets investing. And so what I'm looking for are areas where I can punch above my weight and build conviction in a differentiated view in a efficient way. For specific reasons that only I can capture. And I believe I can capture. My background is unique. I've been able to hold conviction and trade and fets invest effectively in many different geographies in the world. I've not restricted to owning US-based companies. And my experience has been making money in various countries across the world. At the same time, I think the competition has moderated. International investing, I think, is kind of a lost part. And people forget that a lot of the big, famous investment managers, like Julian Robertson, like George Soros, like some of these big groups that have created these heritages over the last 30 years, one that made a lot of money early on in the emerging markets in far-flung geographies. Julian was not buying the large cap consensus long names when he went from one to $10 billion. He was doing a lot more interesting work. But I think over the last 10 years, that they international investing out of the US has become a bit of a lost part. Evaluating management teams in far-flung areas of world, building relationships and holding convictions. This is something that is hard fought. It's hard ones kind of foundational experience and something that I have. The same thing with TMT investing, right? I think people who've made money in Brazil or Mexico or Southeast Asia or India, oftentimes the overlap with technology investing is low. If you look at the indices of these local countries in different places, tech is such a small part. I think tech is less than 5% of the Brazilian index, but it's 40% of the S&P 500. So if you're a local investor, you don't have to know tech. It's not part of your day job. You don't need to build the pattern recognition or the product knowledge to invest and to build, to beat the benchmark in your local economy. How did you come across Nebius and BIS? So I'll start by, I'm a student of history, but I'll start going back to 2015 when I owned Amazon. And there was two things going on that are essentially the same things going on today with Nebius, and that's why I found Nebius as an investment earlier this year. I owned Amazon 2015, and there was a cross geographic reason for a dislocation and a misunderstanding to exist. Amazon at the time, people forget, but they were actually investing a lot of money into China. And at the time I was an investor in e-commerce in China, I owned JD, and I was involved with companies like Alibaba. And Amazon actually at the time was losing almost as much money in China as its entire e-bit line, this is in 2013, 2014. And as the local competitors came up, Amazon started kind of pulling back from China and not burning as much money. That change helped Amazon expand its margins over that period of time. The second thesis I had when I owned Amazon was that there was a significant technology transition going on where this company called Amazon Web Services that was actually quite profitable and growing quickly, and would unlock a pretty significant valuation in the public markets as it started disclosing its growth and profitability. There was a significant board bear debate at the time around AWS, you know, I think from '06 to now, AWS has dropped price over 100 times, right? They've taken price down, and the consistent pushback to that product and that business was, this is a commodity you'll never make money. They disclosed at the first quarter of '15, which I think was surprising to a lot of market observers that the AWS was growing about 50% year-to-year with, I think it was 17%, 18% operating margins, and that kind of changed the dialogue around AWS significantly, right? So, zooming out, there was this geographic misunderstanding where if you understood what was going on in the emerging markets, you had an edge in this global technology business. The second piece was there was this technology shift, where it was hard to see what the long-term value of a business would be today, because the P&L didn't look very pretty. And that was the same as AWS that is today at Nebius. So, let me talk about Nebius. This is a company which came out of Russia's Google, a company called Neandex, it was a public company, and because by virtue of the fact that it was a large U.S. listed public company, it was able to attract some of the best technical talent in Russia, get good mathematicians, computer scientists, logicians, et cetera. So, this is a company that had a dynamic entrepreneurial management team and a founding team, and they had thousands of the best technical talent in the entire Eastern block. So, what happened is after the Ukraine War, the founder kind of came out against it, which compelled him to sell the Google of Russia back to Russia and for a significant discount, right? For one tenth of the value of the business. But he took that money and took 1300 of the best engineers in Russia, and he seeded a number of other businesses. And, you know, having made money in the internet boom, he now saw the AI boom and said, this is we're gonna launch our new business, our next chapter on. So, they kind of came back to the public markets this year and as a new business, which was we have this AI infrastructure as a service business, and we have some investments in a grab bag of companies that many people had never heard of. And, you know, I watched it come back to the public markets. I didn't participate, but I initiated a position after liberation day after it came back down to the realist thing price. You know, my embryo, 25 bucks, roughly, yeah. We, you know, by virtue of having a long and short book and being factor aware and geographically balanced, you know, by April, end of April, you know, we were up close to 10% year date, 9% year date. And so, I felt like I was able to play offense with new positions, new longs that I've been following and covering, right, as part of our sector coverage process for a long time. And so, I dug in and did my work. So, there's two things going on here. One is this cross geographic misunderstanding, right? Russian team, people get scared. They don't wanna think about it, right? When, in fact, it's really a multinational team that's based in Western Europe, in Israel. They have a large presence in other emerging markets as well. And they serve a very global customer base, right? They, you know, it's probably one of the only tech companies in the world that has Tesla andthropic bite dance and mercador lire as their customers, right? So, they are able to serve customers well strives because they're kind of a neutral participant. Not to mention recently, they've signed a large deal with Microsoft. So, there are this kind of, you know, underdog team coming out of a geography that people are not comfortable with. And, but I had a history with, right? I followed this company for a long time. The second piece that's changing here is that they've, you know, obviously are part of this AI transition, right? This AI theme, but what, my original thesis of buying this company was that they had a 28% stake in a database venture that at $25 a share when I entered it, it was, it could be worth the entire value of the company. So, you actually had to put no value on the core business or any of their other stakes to justify buying the company at 25 bucks. And that kind of value unlocked. So, to value that $25 of that piece of the business. Correct. So, that, how did you get to that number? So, I did a lot of work on Click House. You know, I've previously followed the U.S. database sector, MongoDB, U.S. data analytics, companies like Elastic, you know, my previous farm were investors in some of these companies. And I've done a lot of work on the technical capabilities of companies like Snowflake and Databricks, right? Databricks, you know, they've raised private money. You know, I've, you know, been, been, been clued into how they're doing on the private side. And so, the, this is a database product that is significantly faster than any of the companies I just mentioned. That is cheaper. And that when I speak to engineering CTOs of startups in New York, where I'm based in Silicon Valley, you know, I had kind of a panel of these experts and that were my friends. And they said, look, this is a product which is best in class, which is getting better every day. The level of commitment we have and the excitement as a developer and entrepreneur building up business is greater than what I felt what these experts felt like dealing with Snowflake or MongoDB. So that level of a customer sponsorship of this product was something that opened my eyes. I said, okay, you know, there's something real here. This is not a back of the envelope, some of the part story. This is actually a fundamental technology potential winner. And when you think about what makes technology company as good is that you have to have a better product than your competition and you have to have a R&D motion to continue that product lead. And finally, you have to have a go-to-market motion that actually gets you customers and keeps customers. And you know, they, you know, for an emerging markets heritage company, this company, the Clickhouse subsidiary of Nebius, they've brought in experienced, you know, US-based executives on the go-to-market side, on the senior management side to compliment the technical chops of the, you know, Russian heritage team. And, you know, when I look at my universe, I am constantly looking for companies based internationally that have the ability to become global companies, right? And you've seen, you know, over the last 20 years, you've seen a lot of, for instance, Israeli companies, right? Come out and become US tech company, right? That's a common thing you've seen. I'm working on it. That's right, yeah, China, I think, less so because they were such a big domestic market. But right now, you're seeing a lot of Chinese companies go global, right? TikTok is a Chinese company. And people are then to forget that. - A big thank you to Street Account by Faxit for helping in the research process for these episodes. I've been a big fan of this product since it came out many years ago. It's a staple in my process and serves as the top of funnel for all names on my watch list. I just enter the ticker and I'm instantly in the flow for all key news flow, allowing me to ramp up on the name while I manage the rest of my book. And I also use it to monitor customers and competitors for names that I have in the book. They also have value added analysis in their earnings summaries with historical valuation and they report consensus earnings revision, which saves me a lot of time on gauging what level to add to a position or not. Their Gen AI tool, Mercury, is cut in edge and provides access to one of the largest databases for top tier analysis. You should definitely check it out. - Let's go back to the checklist here. - You had a 15, when you bought it, what size, what position weight was in your portfolio? What was it at the peak? Where is it now? And then, you know, is the stock available today at a reasonable price? For sure it was when you were buying it that value. But walk me through how you've thought about the weight and the value, when the stock has gone from 25 to 125, 128. - Yeah, I initially, the position was in the load amid single digits, you know, for four percent, five percent. That's where I tend to start my long positions and sometimes a little lower than that. Amber Road runs a portfolio of 20 to 25 longs and, you know, 25 to 35 shorts. - At peak, it was north of 15%, 15 to 16% of the portfolio. Now, some of the other positions, you know, have gone up as well. I've trimmed, call it 15% of the position, right? Of the, in terms of shares. My price target is $200 and, you know, a lot of folks have talked about some of the parts and I can, you know, walk you through the details of the some of the parts. But I think the upside is, you know, at least in the next year is 200 bucks and I think it's about 30% downside from here based on the most current information. But I want to introduce kind of two alternative valuation frameworks that I look at kind of in concert with a some of the parts. And the one, the first one is a payback to analysis. So, what is, you know, every dollar they put in the ground, what is that, what is that worth? And what is that, what is that earning? And what is the current stock price imply? So if you exclude the value of some of the smaller stakes and that would be a very conservative valuation of those, the, and then you compare to the invested capital and you put a 25 times, 25 times earnings power, you know, a multiple. What I'm seeing here is that the current stock price implies that the market is valuing what they put into the ground as generating a long term a 10% return on invested capital. So, 10 year payback. Now management has said that payback is two and a half to three years, right? Up to the 40% return on invested capital. And if you look, if you look at other companies like Amazon and Snowflake and the legacy software companies like Oracle obviously, it's a lot higher than that 10% return on invested capital. So I think the headline of Nebius's current stock price implies that people think that they are in 10% return on capital. I think, for me, that seems like a reasonable entry point. So how does that translate to what is worth? So let's say you don't believe management and, you know, management saying, you know, two and a half to three years, let's say you think, let's say you say four years, payback, right? A little bit better than there. Actually, they're assuming a four year depreciation, let's say they've returned in four years. The business is valued at 15 times earnings, right? So you put 25 times on that. That's over $200, $202. - Time's what? - Which earnings? You must be looking at that long-term earnings power, right? That return on the payback on the current invest the capital plus their war chest, right? - Well, the long-term, I mean, these guys put out, you know, they're not profitable now. They put out a long-term margin of 20, 30%. They say, if you billion of revenue, how do you discount that tangibly and say, when does that happen? Is it realistic? Do I believe the steps to get to that margin and to that revenue? Obviously, you know, they're, it sounds like they're a really technical management team. They're really good. And there's a lot of demand, right? Obviously, the hyperscalers are putting less on their balance. You're looking for partners. - These are exactly the same questions that were asked about Amazon Web Services in 2014, right? Exactly the same questions. And the path is actually very much similar, right? For a couple of reasons. I think one is, if you look at the biggest debate right now is about kind of the depreciation of the GPUs. First of all, their depreciation schedule is a lot more conservative at four years than core weave at six years. So they're actually putting more on the P&L than their closest peer, which we can get into why I don't think it's a peer. - That's a depreciation though, right? Most of the, when they're doing their calculations, it's cash on cash. - Okay. - And when you're saying they're doing four, core weave's doing six, all the people are doing five to six is, are we talking about the GPU or the entire stack? - That's just the GPU. The entire stack, if you think about the facilities that they're putting into place, the depreciation, the cash headwind from that is much lower than the GPUs. The GPUs are the most financially painful part of the investment outlay. It's supposed to appreciate the most quickly. It will require more capital in the future theoretically. But if you look at what happened to AWS, which they had the same questions, what ended up happening is because demand is so great, you can fill up these data centers even with older chips where people, I think the opportunities for both Nebius and Clickhouse is transitioning from being a source of analytics and inference, right, being where you go to do analytics, inference, to transactions, right, to running your business on these. And where these become trillion dollar businesses is when IBM and JP Morgan and United Health and these types of companies end up using companies like Nebius and Clickhouse for their day-to-day customer transactions and record keeping and CRM and ERP and all that. Going back to the valuation framework and the return on a massive capital, if you look at the commodity part of the business, right, what a core we would do, I would think is almost like a real estate business. You're renting out capacity. And at early days of Amazon is the same thing, compute storage and bandwidth. And we just buy upfront and rent it to our customers who don't want to do it themselves. Let's say that piece of the business is break even. Now for AWS and for Nebius over time, a part of their revenue is going to be higher value. They're selling analytics, databases, machine learning operations. If you look at Amazon, right, they've over the last 15 years they've come out with Redshift and Aurora and other higher value added products, not just selling terabytes and bandwidth and compute hours. And if you assume that a piece of the revenue, even if their current fully paid for capacity and you look at what that kind of incremental margin should be, what I calculate is that software stream at the current Nebius valuation is trading at 25 times earnings, right? That software stream. Assuming the core infrastructure business makes no money, right? Now you compare that to a snowflake trading at 150 times plus earnings. That's an extremely example, but 25 times earnings for a pure software business that we'll be growing in the future is too cheap, right? So that's a very onerous way of also looking at the earnings power of the business, even if you don't believe that the commodity business ought to make money. So that's another alternative framework, differently presented than a traditional sum of the parts or even the payback analysis that we've looked at before. So going back to the payback analysis, you put 25 times this 25% return on invested capital. And by the way, this is only into the war chest they have today, right? We're not assuming any new capital. There's no assumption, may no value accrue to the platform nature of this business. The fact that you can raise capital and apply creatively, right? I think once people start thinking about that for any business, it becomes a little, the evaluation becomes a little high. In Amazon's example, when people start putting valuations on future businesses, that they haven't even talked about. When I remember every time that Ali Baba people started putting a sum of the parts value for this enormous international US business that they were going to have, that was usually a clue to exit the stock or to short the stock. In this case, no one's talking about the platform value of the business, which is real, right? You and I, right? We could not go out and issue equity or cheap debt and start a AI infrastructure as a service business, but there's a small cohort of companies in the world they can. We don't have to value that, right? All we have looking at is the $5 billion capital and they've already put in the ground and the $4 or $5 billion of dry power they have, assuming a 25% return on capital on that business, trading at 15 times that earnings power, 7% yield. I don't think that doesn't make sense to me. You don't have to put 150 times, snowflake earnings power, snowflake earnings multiple to get a lot of upside. So I price targets around $200. Downside is about 30% downside, kind of $80, $85. My sizing has come down primarily driven by my lost budget. To your point, this is a company that has been a multi-bagger in a six-month period. The volatility of the stock is high. And so for a 15% position to go down 25%, 30%, that's a significant P&L hit for a guy like me building a business. But I think that what keeps me invested, other than the reasonable valuation framework that I presented as well as the rich catalyst path, which I can talk about, is the portfolio construction, right? I have AI thematic stocks on the short side that ballast out the thematic ups and downs of AI. And I have, as I described earlier, kind of value event-driven stocks on the long side within international technology that provide a different factor profile on the long side as well. - Today you still see a good risk we were with $75 upside and call it $35 of downside. - Yeah, that's right. - Can we talk about the catalyst path that you see going forward? - Let me start with a negative catalyst path, right? So one thing I like to do with all my companies is do a pre-mortem. If this didn't work, why didn't it work? So let me start with a negative catalyst path. And the first negative headline, we'll go back to AWS as an example, right? Keep going back to that because it was such a successful example. AWS cut price over 100 times. I think the average price of the products is down 15% year-to-year. Certainly, even more so 10 years ago. And I would suspect that you will see headlines about pricing compression in AI infrastructure as a service at some point, right? Right now, there's a supply shortage, right? So prices are pulling steady. And in some cases, they're going up. That's not going to last, right? You're talking about new rental prices. - Okay, you're talking. - Unit prices, exact unit prices. In order to achieve widespread adoption, you got to see these prices come down. For inference and for training, you've got to see prices come down and they will come down with increased scale. You're talking about the rental price per hour, but Jensen's going to talk about the number of tokens price coming down and not the rental price per hour. How do you square that where they can still get a good, healthy rental price, but the number of tokens that you're paying for as a user is going down. So that's actually what you need for adoption. - Well, so I think that's where some of the positive catalysts on the nebius side will come in. And let me look, I will get to that. And they center around the software layer that nebius is building. There are other negative catalysts that will come out. For example, you'll see some of their peers, like CoreWeave, have their own ups and downs, videos syncratically, and nebius will get pulled in with how CoreWeave trades. I do believe they're different businesses. They're different businesses in terms of quality, in terms of their loan through strategy and in terms of their culture. CoreWeave was started by hedge fund guys, right? And they were started by energy traders in Jersey City. And they, I think their approach is very much more about a real estate type of approach, right? We buy them, we rent. Whereas, nebius came from Yandex, which is a technology company. They're started by technologists, software developers. They think the way that CoreWeave trades and the way that Oracle trades and the news flow around, indeed, GPUs and AI will impact nebius' trading. That is part of the negative catalyst path. The third piece is, one day people might wake up and say, oh, wait, this is a Russian team. And they sell to Middle East and sell to China. They have a global customer base, which they're able to because they're not an American company. But people might not like that, right? And there will be political ramifications to having this kind of global business where you can serve a diversity of customer. And you're not aligned with any single GPU political access. But as an emerging markets investor, you think about that, right? And thankfully, I have shorts that also are potentially aligned differently than the current US administration would want you to be. So that's part and parcel of investing in the emerging markets. And it's just part of my day job is navigating that. So let's turn to the positive catalysts. The biggest positive catalysts that I'm looking for start with the product side. New product launches and new integrations. Similar to how Amazon, over the last 15-- they started in '06, right? So almost 20 years now, some of how they built the business. They went from commodity, commodity rental of compute storage and bandwidth. And they went higher and higher up the chain. I think at one point, Amazon had a customer relationship software product that hurt sales first stock for a while. But that's something I would like to see. It's like in pieces, click house fundraising, right? They recently raised about $6 billion. I believe the company could probably raise its next round at north of $20 billion, just based based on the speed of growth, based on the valuation of some of the public cops in the area and the grade of talent and VC relationships, the ecosystem that click house is part of. Now part of that can be secondary, right? Maybe it's made sure to exit some of that and take a couple of billion dollars off the table. But I believe that the next round would likely be primary. The opportunity is just so big for a best-in-class data warehouse database product like click house. The third positive catalyst is new partnerships. And this year, they've signed a very large revenue agreement with Microsoft. The stock was up 40% on the back of that headline, 19 billion revenue deal. There will be others. I can't speak to what they are. But there will be others because they've got technology leadership. They are a neutral party in the AI game of France. And so they have the ability to serve customers all of the world. And I would be shocked if there weren't new partnership announcements in the next quarter. And then I think finally is more of a technical catalyst where you will see over time a company this size get index inclusion, get institutional sponsorship. So a couple parts of that. Let's just start with the last one on being included in indexes. They are an emerging markets company. So they would not be in US indexes. They would be in international ETFs and such. It's a much smaller passive group unless they move to New York or something. Fair, but they very well cook. There is a well-trod path for managing teams that have launched their businesses in different places and made their way to the US. I'm a customer of a buy-coding company. It's a private company called Replet. The managing team, the founders, they're Jordanian. And they now, their team is in the Bay Area in California. So tech companies find their way here one way or another. And when you talk about big contracts with partners, are you talking about hyperscalers trying to get bare metal contracts off their balance sheet? Or are you talking about someone like a Tesla or an actual corporate customer using them for something like an AWS alternative? More of the former, they already have these big customer relationships. And I think you'll see, not just, I would call it, hyperscaler bare metal partnerships, but other partnerships that could make sense are the very large-scale software players today. Oracle is the one that everyone thinks about. They kind of led the way on that. But they are trying to be a hyperscaler. But I could see somebody like Anebius partnering with companies like Salesforce or other end user-focused facing software companies. I could see them making a partnership with Amazon. Amazon doesn't have these types of relationships today. And what better partnership for a management team that says that they want to be the next AWS, then potentially with AWS. So the University of Partners is pretty broad out there. But right now we're in a shortage capacity for GPUs and compute. But that'll go through cycles. Why does someone come to them versus go into an AWS or a Microsoft or even a Google? What are these advantages that you're talking about on, I guess, the technology or performance side? Yeah. So this is a product that was created by technologists, not by people who are thinking about renting out space, or renting out capacity. And so when I speak to the startups that are using their services, the feedback tends to be-- just from a qualitative perspective, feedback tends to be faster, more scalable, and lower error rate. And this is across their different products, across Anebius and Clickhouse. In some cases, the pricing is lower. In some cases, it's comparable. Pricing is not a huge differentiator. It's more about the speed, the scalability, and the ease of access for the smaller customers. And that, I think, is where you really make high margins. It's with the startups, with medium-sized businesses, with growth teams at larger companies. I'm not talking about the bare metal customers. I'm talking about customers who want to use the full stack. They have products that will help you leverage the open source language models. And the ones that are coming out of China, and they're coming out of Meda, or Lama, or QN. They are integrated well with these models. And they're essentially selling a free product, but not for free. And so that is a very high margin product. And people love that, because it's turnkey. It's all in one. It's a one-stop shop for whatever they want to do with AI. And a lot of companies right now are kind of in the experimental phase, which I think is what's really exciting. It's experimental phase for this entire sector. The workloads that are running through Clickhouse are really partially experimental workloads. And so for them to go to production, and then for them to go to the transactional stage, and that's where it's going to get really exciting. Where you and I are interacting with AI, not anthropic, but just organically through a voice interface when we're trying to make an appointment with a doctor, or when the hedge funds risk management system services unknown or unproceeds factor exposure recommends trades to utilize that factor exposure. That we're still just getting there. We're not even there yet. And so that's where you really see this turnkey, higher value and add way of selling AI compute storage in bandwidth. That's where you really see it come to play. I'll have to admit, I was using Lovable Vine code in last night. So even I'm getting up on being a user of the AI code. I want to touch a little bit more on the other side of the bet. And you said, we're in a shortage situation. They're getting good rental rates right now. You've kind of mentioned the customers are experimenting. How do you view the returns the customers are getting from using their services? What are their ROI's? I think the ROI is incredibly high, because there are a lot of successful cases of-- I'm going to bring up case studies in the emerging markets of companies that have taken AI and machine learning. And they've been done doing it for a while. And it is a core part of the business. There are digital banks in Latin America, in Southeast Asia, consumer finance apps that use large language models to determine who gets credit, and I want price. And the way they do that is based on integrating multiple sources of information, for instance, location data, voice, app metadata, device metadata, credit histories, transactional histories from e-commerce platforms. And they're doing all that with AI. And to a certain extent, international markets are more open to using AI for something like credit scoring. I think there are certain regulatory barriers in the US. There's a FinTech in Indonesia that I'm invested in, where when you call them to negotiate your delinquent loan based on your tone and how you speak and your vocabulary, they can determine what your repayment history could be. And that company happens to use some open source Chinese large language models. Inter and new bank and Mercado Libre in Latin America do the same thing at a bigger scale. This is great, man. I've learned a lot. I appreciate you coming on. Are there any final words you want to tell us about the Nebius thesis in Amber Road? Yeah, I think that Nebius example of the type of opportunity that I believe is uniquely available to Amber Road, where we have a massive catalog of case studies driving our patent recognition for technology companies globally going through a period of transition. We also look for opportunities where there is cross geographic confusion and misunderstanding. Local investors, not understanding technology or US investors who don't understand individual kind of international markets. A lot of companies today are becoming multi-nationals. A lot of Chinese companies are going to Asia. Latin American companies are going to the US. European companies are going to Imiya, broadly. US companies are always multinational. So the confusion geographically just creates opportunities for us. And then finally, we're going through this technology transition today, right? AI is a mega trend. But it's hard to research. It's hard to gain commission when things are changing so quickly. You need a process and you need patent recognition. I remember pitching as an analyst in 2014, Microsoft, which had just released this product called Office 365. It was their first cloud product. And they were going through the transition at the time that Adobe had gone through to go to cloud. And people were confused. Why would you sell this thing as a subscription? But so there was this transition going on in technology, which is difficult to understand and difficult to gain conviction. But that's where there is the opportunity to really generate a lot of value as a active stock picker in today's market. One last question for you. What in-- do you think we are in in the AI CapEx boom? AI CapEx boom. I'd say we're in ining two to three. OK. We did a survey on LinkedIn. And we got about the fourth in. So pretty close. I think it's because my perspective being international, right? International is always earlier, right? It's always earlier. But it always tends to move almost faster. The two biggest things that you see in emerging markets are it starts from behind. But it can leapfrog very quickly, because they don't need to go through the intermediate steps. The startup today in the emerging markets, who's looking for their tech stack, is not going to think about the buying servers. He may not even think about Amazon Web Services. He might jump straight to a large language model enabled AI infrastructure as a service like Navius, right? And so that's where I think maybe my perspective is different. I think there's a lot more to go. And a lot of large businesses in the internationally, they still run their businesses on paper. And you really can't say that for the US anymore, right? And so I think that for setting where I am here, there's at least 10 years of rapid growth in these different technology sub-setters I cover in the countries that I'm exposed to. What does this mean for the current hyperscalers? If you have these neoclouds that are a bit of leapfrog, the hyperscalers are trying to do both, but are they going to get left behind or are they going to be able to have a competitive offering? Two points here. One is that the hyperscalers will partner with the neoclouds and they have, right? And the second point is that there's room for both to exist. The great thing about software and about cloud is that they tend to be pretty sticky. Once you have your production workloads and your transactional database in a certain environment, you tend to stay there. But where there's disruption is, tends to be on the net new, right? The new companies, the new business models, the companies that are going after the competitors to the customers of Amazon Web Services, right? The US Regional Bank will be on Amazon Web Services after a 10-year digital transformation effort. Great, and they're never going to leave. But the digital consumer wallet that's going after the Regional Bank and the US stock guy will build from the scope of scratch on a new tech stack, right? The tech stack 3.0. And I've seen this through every technology transition. What happens is the incumbent businesses are on the incumbent stack and they all do fine, right? They all grow, but then your net new capital, your net new businesses, they get built on a new tech stack. And there's room for everyone. Oracle has been around for many years, right? And they've managed to transition through these different technology shifts. I have no doubt that Amazon Google will do the same. But the economies grow, right? There will be room for new players. And I think it's relatively healthy to have that dynamic. Well, thank you for coming on. I have to remind our listeners that nothing on the show is investment advice. And the stock might not be suitable for you, and you should check with your financial advisors. Please follow me on LinkedIn and give us a like on your podcast channel. If you liked it, it goes a long way for us and feel free to reach out to me or synon. Thanks, everyone. Thanks, Doug. Take care. Just make sure to follow "Pitch to PM" on your favorite podcast app and never miss out on episodes. This show is for entertainment purposes only. Please consult a professional before making any decisions. This show is copyrighted by "Pitch to PM." Written prevention must be granted before re-bark fasting.