Why Anthropic Meta And Tesla All Chose The Same Database Aaron Katz Clickhouse

TITLE: Why Anthropic, Meta, and Tesla All Chose the Same Database | Aaron Katz, ClickHouse CHANNEL: Weights & Biases DATE: 2026-03-31 ---TRANSCRIPT--- in August of 2021 with a $50 seed round, possibly the largest pre-seed round in the history of enterprise software. Because I didn’t have a pitch deck, we had no product, we had no [music] customers, and we had no revenue. Microsoft, Siebel, Oracle, they basically were on record saying cloud [music] computing is a fad. No one’s going to move their data to the cloud. We were on this mission to create this category and refute all of these baseless claims that it wasn’t going to work. ClickHouse is a database, satisfies a ton of different use cases, [music] and you look at it as a platform potentially versus various applications or solutions. [music] We have over 3,000 customers on ClickHouse Cloud, we add hundreds every month. And Anthropic asked Claude what they should be using for a specific use case, and Claude suggested ClickHouse. I called my head of Europe, Arnaud, and I said, “Wake up the bankers at our bank in Netherlands cuz there’s a $100 million wire that’s coming through [music] in the next 30 minutes, and it needs to clear.” And so, we wired $100 million, it cleared, 3 minutes later SVB’s banking [music] system went down. You’re listening to Gradient Descent, a show about making machine learning work in the real world, and I’m your host, Lucas Bewald. This is a conversation with Aaron Katz, who is the CEO of ClickHouse, one of the fastest-growing databases of all time, and a company that I have long admired. This is a fun chance for me to talk to someone who I’ve learned a lot from, and I get to ask him about how he thinks about his personal development and growth, who he admires as a CEO, and also we talk about his recent acquisition of one of our competitors, and whether or not he plans to compete with us. Hope you enjoy this one. All right. Well, Aaron, when we first met, you told me the story of ClickHouse and how you got involved, and I remember it was one of the most fascinating stories of the genesis of any startup, and tell me the whole story. All right. Well, I remember you and I first meeting down in Santa Barbara, and we went on a a walk along the beach, and and I kind of told each other our respective founding stories. It feels like that was yesterday, but I think it was probably 4 years ago. Um and so, uh ClickHouse was developed by uh one of my co-founders, Alexey Milovidov, and he developed it when he was working at Yandex. Um and he uh was looking for a database that would be able to ingest and store petabytes of uh streaming data from Yandex’s various services for web analytics, and nothing off-the-shelf was able to scale and provide the performance that uh was needed, and so he developed uh ClickHouse. It’s short for clickstream data warehouse. So, he was thinking about the data warehouse use case uh back in 2009 when he uh developed the the database. And it just took off inside of Yandex for a a variety of different use cases, and he had the uh the brave courage to convince the Yandex leadership team to open source it in uh 2016. So, 7 years later, they had formed a team around ClickHouse. Uh it was one of the fastest-growing uh databases uh after it was open source in terms of number of contributors um and number of uh of contributions, and it still is. And that’s when I first discovered it. When I was at a previous company, uh we started to bump into ClickHouse in the market. We actually started to lose a lot of uh observability workloads specifically uh to ClickHouse. Uber was migrating their logging infrastructure to ClickHouse. We lost a big metrics project at eBay. Uh Disney, Comcast, Deutsche Bank, and others were were moving to it. So, it was on my radar, uh but there were a lot of uh you know, popular open source OLAP engines at the time. Again, this is 10 years ago. And uh so, during COVID, I I stepped out and started thinking about what I was going to do next, and uh an investor asked if I’d be interested in talking to the Yandex uh leadership team about ClickHouse, and I I just jumped on it. And this was early 2021. And uh I kind of you know, in the thick of COVID, and so everything was done virtually at the time. And so, I you know, I met the Yandex leadership team and and Alexey, and uh they you know, they said, “You know, we realize that ClickHouse is very popular, and uh we think that it’s got a lot of potential, and we’d be curious to know if you were interested in in doing something with it.” And I said, “Absolutely. I would I would love to partner with Alexey and form a company uh around it, and uh you know, it’s going to have to look and smell like a Silicon Valley startup uh with you know, a a venture-backed uh Delaware corp, and this is something I I think we could really um un- unleash the potential of this technology by building a fully managed service, um a serverless offering with compute storage separation that could automatically scale and idle because these analytical workloads are are very bursty. And we could take a lot of these companies that are running this technology themselves either in the cloud or on prem, and give them all of these amazing benefits of a hosted offering called ClickHouse Cloud. And so, I reached out to two uh venture capitalists who I’d known in the industry, uh Mike Volpi, at the time he was at Index Ventures, and Peter Fenton at Benchmark. And in the world of open source infrastructure, Lucas, as you know better than anyone, there’s really a handful of of investors who you’d want to partner up with and have experience with these business models and understanding all of the nuances around open source. And so, it took us about 9 months to pull off because at the time Yandex was a publicly traded company, I think with a $30 billion market cap. And so, when you take intellectual property out of a public company, it comes with a lot of uh regulatory concern and consideration. And so, there were a lot of lawyers involved as you could you could imagine. And I I uh I convinced a uh an executive at Google named Yuri Izrailevsky, and I’d met Yuri when he was at Netflix, and he ran platform engineering at Netflix. At the time, they were AWS’s largest customer, I believe, and he he really led their migration to the cloud, and he’d been building distributed systems around open source for 20 years. And I kind of convinced him along along with others to uh leave Google and and join Alexey and I to form the company. And so, uh three co-founders with uh Yuri running product and engineering, Alexey as the creator of ClickHouse as our chief technology officer, and then myself as the CEO. And uh we got it done, we pulled it off in August of 2021 with a $50 million seed round. Maybe it was even a pre-seed round, possibly the largest pre-seed round in the history of enterprise software. Because I didn’t have a pitch deck, we had no uh product, we had no customers, and we had no revenue. And uh but it was peak ZIRP period, and so there was a lot of capital getting put to work. And then we quickly followed it up with another $250 million round that was jointly led by Coatue, uh where you and I met, and uh and Altimeter. And there’s a lot more to the story, but that’s kind of the origin story of ClickHouse Inc., the company. And you’re looking at the headquarters here in Silicon Valley. Um and Alexey and his team are in Amsterdam. We’ve got a really strong engineering hub uh in the Netherlands. Uh you know, where are we 5 years later from when we formed the company? And when you were doing this initial pre-seed round that was absolutely enormous at that time, people might not realize how what an outlier it was at that time because pre-seeds have gotten so astonishingly big in in some AI sectors. What when that was happening, what was it that gave you and your investors the confidence to do that if there was no revenue and not even a product launched yet? Well, we had thousands of companies using ClickHouse, the open source distribution, for a wide array of use cases from uh analyzing clickstream data to data warehousing, to observability, as a cyber back end. I mean, I think Microsoft is on record that some of their largest analytical workloads run on ClickHouse, like Microsoft Titan and Microsoft Clarity. I mentioned some of the other companies that were migrating to ClickHouse. Um and so, we had a a lot of conviction that the technology itself was highly differentiated to start with. Um and so, I feel like we had a bit of a head start, frankly, um than most startups uh because we had this very feature-rich database, and we had the core team of committers. As you know, in open source, you can have hundreds or thousands of contributors, but you know, to really uh control the road map and the direction of the project, you really need to employ the committers. And we’ve seen in the past what happens if if an open source company does not do that. And so, we felt like we had the right team, uh Yuri, you know, having such success building distributed systems around open source databases, myself on the distribution side, you know, having studied go-to-market for 20-plus years at at companies like Salesforce and and others. And so, I felt like we had the right team assembled, and I think uh the investors agreed with that thesis. We had a a great group of companies as design partners early. We had 50 companies that were using open source ClickHouse saying, “These are the characteristics of a cloud service that would compel us to move from self-managing the database to a fully hosted service.” So, we had a lot of input from early adopters of ClickHouse. And then we just saw a ton of market pull. Um as you know, because we cover such a broad surface area, there’s a lot of competitive dynamics that we need to navigate. On the data warehousing side, you’ve got Snowflake, Redshift, BigQuery, et cetera. On observability, it’s a very crowded market, as you know. Um but really ClickHouse shines in terms of low latency um analytics uh that are customer-facing, like the Weights & Biases weave product, uh which is built on uh ClickHouse Cloud. And so, you know, we we felt like we had a huge TAM to go after. We had the right team to develop a differentiated service. And then many of us had had been part of open-source companies in the past. And so, we kind of you know, had the unfortunate experience of knowing what to avoid. Um because getting getting it right with open-source, as you know, is very tricky. Only a handful of companies have been able to to achieve it. And that’s why we prioritized this cloud service uh before everything else. I actually wanted to ask you. I saw that your father worked at Xerox PARC in the really early days of Silicon Valley, I believe. And I wonder how growing up in that environment might have shaped your perspective on technology. Well, he worked initially at Xerox out of college in the ’60s selling copy machines in the Bay Area. And I think his largest customer was Stanford University. And this is when Xerox, really Xerox and IBM I think were widely recognized as the most innovative technology companies uh in the world at the time. And he then went to Xerox Document Systems, which was co-located at PARC, which for those listening stands for Palo Alto Research Center. And is where a lot of the innovations that we use every day were created and invented. Things like the graphical user interface, the concept of the mouse. Um it was just revolutionary. And so, I was a kid, Lucas. Honestly, I was you know, I just turned 50 yesterday. So, I was you know, born in ‘76. Um and so, we’re talking about the mid-’80s uh when in the period that we’re describing when my dad was was working at Xerox Document Systems maybe the the late ’80s. So, obviously, I wasn’t paying too much attention to it all. I think there’s the benefit of just kind of osmosis and being around it. Um but I do remember going to PARC uh with my dad and sitting in his office and doing homework or reading a book or doing you know, LEGOs or God knows what. And so, I think just being around it and and seeing the energy um was very appealing to me. So, before I even started at college, I went to UC Davis, it was clear to me I was going to go into technology. Like I grew up around it. I saw the lifestyle that it provided. I saw the joy that my dad got by building and leading teams and securing a a customer win or helping a customer solve a problem. I don’t know. For me, it was just the same way that if you grow up and your mother or your father’s a doctor, there’s probably a decent chance that may be a path you pursue cuz you’re just around it. You live in it. Totally. Um and then you went to Salesforce, I think, in the first .com uh or the to what to me is the first .com uh bubble um era. I was wondering if that if that kind of shaped you. And I was wondering if seeing that big bubble gives you a perspective on the possible AI bubble that we’re in now. Well, of co- of course it shaped me. It was transformational. I mean, Salesforce the 12 years I spent at Salesforce were so pivotal in my life for so many different reasons. Um and the other reason I went to Salesforce back in early 2002, so this is 24 years ago, was because I didn’t get into business school. The plan was to go get my MBA. And coming out of the dot-com bust in ‘01 and early ‘02, I applied to a handful of I think what many would perceive to be elite business schools. It’s Harvard, Stanford, etc. And I didn’t get in any of them. I got waitlisted at MIT at Sloan. And so, I was unemployed because the startup that I was working at had failed. And I was living up in Lake Tahoe waiting tables. And so, I needed a job, frankly. And so, I came back to San Francisco and I was able to navigate the interview process to meet with Marc Benioff. And I remember that interview vividly cuz he asked me one question. He said, “Why should I hire you?” And somehow I convinced him that it was a good decision. And at the time, it was a 150-person startup. It was a 30-person

How did you answer that question? I I I got to know. I said, “Because I’m going to I’m going to outwork and be better than everyone else. And I’m going to be your top-performing sales I was a sales engineer, which now people call solution architects. Maybe they call them forward-deployed engineers. I don’t know. But you get it. Technical pre-sales. I cuz I’m going to work harder and I’m going to be smarter and I’m going to out-execute my peers. And I knew that he was very competitive cuz he was a disciple of Larry Ellison at Oracle. And so, I knew that he would I think respond favorably to somebody that had a competitive mindset. Um And I’d gone through the gauntlet of interviews to get to him. So, I’m sure I suspect I wouldn’t have gotten to him if the team didn’t think I was going to be uh a good hire. Nevertheless, um how does that correlate to the AI bubble that we’re in today? Well, I I do think there are a lot of parallels. Cuz Salesforce was, you know, really like software as a service was not a new thing. It had been tried and failed during the dot-com bust. But Salesforce was the one that made it a reality and created the category and it spawned so many other companies. ServiceNow, Workday, we could go on and on. Um and so, at the time, if you recall, like our competitors Siebel, SAP, Oracle, Microsoft, they basically were on record saying cloud computing is a fad. Like no one’s going to move their data to This is before AWS. No one’s going to move their customer data to the cloud. It’s not secure. It’s not reliable. It’s not durable. And so, we were on this mission to create this category and refute all of these baseless claims that it wasn’t going to work. That all software was going to be run in data centers on prem. And so, if you apply that to the current AI themes where people say, you know, it’s we’re we’re never going to realize super intelligence. AI models are unpredictable. Uh you cannot trust them. Uh you don’t want to send your enterprise data to an LLM provider because it’s your competitive edge. All these things are going to be debunked if they’re not already. And we’re all going to be using AI the same way that we all use cloud services every day. Even in large regulated industries, they’re still using cloud services in the in the company. So, I see AI as as no different than, you know, early skeptics or incumbents who are have their head in the sand. And they just like don’t want to accept the reality. The same way that Siebel didn’t want to accept the reality that their core business was frankly And they ultimately got acquired by Oracle. And it was a it was a good company. Salesforce you know, replicated a lot of the Siebel feature set um from a CRM standpoint. But there was this this paralysis that set in that they were unwilling to accept that cloud computing was going to be the way enterprise software was delivered in the future. And I think you you can either accept the reality that we’re all going to be using AI models much more than we are today in a much faster period of time. It’s not going to take It took us I joined in ‘02. The company went public in ‘04. We still only had a handful of enterprise customers when Salesforce went public. Like AI is getting adopted at a 100 times faster than software as a service was 20 years ago. When you were creating the go-to-market motion for ClickHouse, [snorts] it seems like you built it in a really modern way. One of the things that I really admired about ClickHouse is when we bought it at Weights & Biases, it really felt like you guys met us where we were at. It felt like a very technical sale to a technical audience. There wasn’t a lot of pomp and circumstance. Nobody was asking us to go play golf or get dinner. It was just fixing our problems, getting us up and running, waiting to charge us until we were up and running, and being incredibly um effective at giving us timely technical support. I really admired that. I I I went to my team and said, you know, we got to learn from this this ClickHouse team. But it’s interesting to learn about your background. You you have this kind of you’re probably deeper in Silicon Valley go-to-market motions than anyone. Your dad was doing it in in the ’80s. What do you think you like took from that early experience with Salesforce and before that? And and what did you kind of leave behind as you were building out the ClickHouse go-to-market team and motion? Well, when I was starting the company, as I mentioned, the strategy was going to be to take this very popular open-source database and build a managed service. And I looked at the other open-source companies at the time who had gone down this path. And from where I sat, MongoDB was the best of the bunch with their Atlas service. It was a broad database platform that satisfied a lot of different use cases. So, that was kind of the first design principle. But that doesn’t mean that you’re going to be successful in building a service that distributes itself to a highly technical audience. So, then I looked at the next two proprietary infrastructure companies at the time, Datadog and Snowflake, about how they went to market. And they were very different. They were both very successful, but through different means. Datadog built this offering that a developer could evaluate, deploy, scale in a frictionless way. People call it PLG now. Like it’s some sort of new thing. I mean, at Salesforce 20 plus years ago, we had a free trial. Yeah, I was that PLG? Of course. But in terms of infrastructure, the goal was to develop a service that you or another engineer could stand up, evaluate, test, even push into production without ever talking to somebody in sales. And if they needed to talk to somebody in sales, they weren’t going to talk to a salesperson. They were going to talk to an engineer because they wanted a peer discussion. And then nobody wants to be sold to. People like buying things, but nobody likes being sold to. And Snowflake, by comparison, invested heavily in sales and marketing targeting the enterprise buyer. They were very successful in doing so. It took them a lot longer, and it required a lot more capital. I just thought the Datadog path was going to be a lot easier, frankly, and a in a faster time to market. And so that was kind of the design principle behind ClickHouse Cloud. And then when somebody like you or somebody at LangChain or Vercel or Lovable or Decagon or Sierra or Anthropic or OpenAI or anybody else that’s using ClickHouse, when they wanted help, they weren’t going to get sold to. Like we were going to establish a joint Slack channel with our top engineers, including Alexey and our committers, and we were going to make them successful in such a short period of time that you’re going to be in production before our competitors have even scoped the project. And so we weren’t going to charge migration services. Let’s say you’re migrating off of Postgres, or you’re migrating off of Snowflake or BigQuery. We were going to help you do that without any investment for in terms of your capital. You’re investing your time to to evaluate our service. We were going to make you successful as soon as we possibly could. And that was kind of the spirit, and 4 years later, it still seems to be working. Another topic that I love to ask you about, which is the SaaS apocalypse or the the cratering of of SaaS company valuations, which on one side seems very extreme and has hurt a lot of good friends of mine. On the other hand, even at Weights & Biases, we see more of our customers engaging with our products like a database, using agents to log into our software instead of humans. I’m curious where you stand on that. Well, you know, one of the few benefits of age is experience, since I’ve been through a lot of cycles uh in our industry from 9/11 to the financial crisis of ‘08 and ‘09 to the SaaS sell sell-off in COVID, etc. So like, this is I mean, this is significant. I don’t think we’re going to get back to the multiples that uh the people have enjoyed in the past. So I think this is a new normal uh in terms of uh what companies are going to trade at. Um I think it’s a little overblown, and I think there are companies like Datadog and Snowflake that are being incorrectly associated with SaaS companies like Salesforce and ServiceNow and and Workday. I I think they’re very different. Um I do think So valuations go up and down. Like you know, I try to tell our employees that all the time. Like I’ve seen companies where employees get so fixated on the stock value and evaluation and the stock price that they get they get drunk on it, and the hangover is rough when all of a sudden the valuation gets cut in half, and you’ve got to have a long view. Um not everything is always up into the right. Um it’s been it’s been great for ClickHouse over the last 4 years, but we will be tested and tried like every great company over the course of decades. And that’s kind of the time horizon you have to have in mind. Um I do think serving at the infrastructure layer is a very good place to be right now. Um and if you simplified it, it’s kind of picks and shovels to the gold rush, um which is uh everyone’s going to need a database. And what are the characteristics of that database? It always comes down to price and performance. Obviously, when you have a hosted service, reliability is P0, durability, security, scalability, all those things. Um but when somebody makes the purchasing decision, it’s going to come down to price and performance, assuming everything else is table stakes. And so um I don’t think we’re going to get back to the levels that we all enjoyed in the past from a multiple basis. You just call it a multiple on forward revenue. Um I think these companies are being valued on their free cash flow dynamics more so today than they ever have been in the past and the terminal value of the company in the future um more so than just expecting that these companies are going to generate 30 to 50% year-over-year uh year-over-year growth. Um I think it’s a really good opportunity for companies like ours that are well capitalized and private, and we’re seeing that with Databricks, for example. I mean, they’re reaccelerating growth on a $4 billion uh revenue base across a variety of different product lines. I think their data warehousing product offering is past a billion dollars of revenue. Their AI-related workloads now is over a billion dollars of revenue, and they’re accelerating growth. So I think Databricks, ClickHouse, companies that are designing for agents, not humans, are going to get a lot of lift uh over the next end number of years. Um and I’m thinking about a world, you know, Anthropic presented at our user conference last year with you, Lucas, when you were on stage talking about why you chose ClickHouse. Anthropic and OpenAI and Tesla were also uh on stage presenting. And Anthropic at the time asked Claude what they should be using for a specific use case, and Claude suggested ClickHouse. And then I was talking to the CEO of one of the largest fintech companies in Europe uh a few weeks ago, and he said, “Every LLM I ask what I should be using to re-platform our company suggests ClickHouse.” Which is great. So but but it still requires a human to make that request. I’m thinking about a world where these agents are actually selecting and provisioning the infrastructure behind an application. So you you build me an application that needs to observe telemetry, whatever. For the and it’s it’s going to go and it’s actually going to not just recommend ClickHouse, but it’s going to provision a service, uh and it’s going to stand up the stack, uh and it’s going to not only have an analytical database, but they’re they’ve got transactions that they need to store. So they’re going to provision a Postgres service. And we’re developing a managed Postgres service like many, but we’ve been thinking about this for years. It’s in private preview today. It’ll be in public beta uh in a few months, and it’ll be generally available by the end of the year. And we’re going to have this unified data stack uh where you can have both an analytical uh workload run on ClickHouse, and you can have a transactional experience built on Postgres. And it’s not going to be yet another managed Postgres service. It’s going to be the world’s fastest Postgres service in the cloud. But I guess when I look at Datadog, I’m also a fan. Been I’ve been a fan for a long time, and I I admire the team over there, but that seems like a harder position, honestly, doesn’t it? I mean, they sit there at this layer where developers love them because developers are logging in and enjoying the thoughtful interfaces that they’ve built. But if it’s mostly agents logging in, that seems like a massive disruption to their business model. And I do see more and more people just using ClickHouse directly as their observability layer rather than going through something like Datadog. And now you can easily build these custom interfaces for your particular application. Well, I I I think that’s a valid uh way to look at it. And it’s it’s as you know, it’s impossible to look out 3 to 5 years right now in our industry. It’s changing so it’s changing faster than I’ve ever experienced in my career. So I do think if you look at, you know, 1 to 3 year time horizon, which is really the only realistic outlook, you can see Datadog still being a very meaningful supplier in the industry. It’s got it’s such a premium product. Uh people really underestimate uh how difficult it is to compete with Datadog um and develop a comparable service. We tried at my previous company, and I would not say we were successful. So I think we have uh a reasonable view on how formidable of a competitor Datadog is and how high quality of a product it is. I’ll give you an example. Lucas, we we used Datadog uh internally used in the past. When we were developing our service, it was the obvious choice uh because we had this incredibly compressed time frame to launch ClickHouse Cloud. And standing up a Datadog service for APM and logging and metrics was was the obvious choice. And it grew like wildfire, like it does in many companies. And all of a sudden, I’m spending seven figures this before I had any revenue on Datadog. Like there was a period where I was spending more on Datadog than I was generating in revenue as a company. And we also knew that you could use ClickHouse for observability, like you just mentioned. We didn’t have HyperDX yet, so you had to rely on uh something like Grafana, um which is an another great product as an alternative to Datadog. And so we went to the company, and we were at the time about 100 people. We’re now about

500. And we said, “Hey, we got to migrate off Datadog.” Like we got to use our own technology. A, it’s too expensive. Um B, you know, the concept of dogfooding, drink your own champagne, whatever you want to call it, is really important. Uh we can’t say that we’re a meaningful supplier in observability if we don’t even use our own technology. And it it was almost like we had a mutiny in the company. Like our developers like you you can’t take data It’s like I was taking heroin away from a drug addict. They’re like you you you can’t take my DataDog. I’m like we have to migrate off of DataDog. Like there’s no question we need to do this. It’s just a function of how quickly. And it took us longer than I would have liked and eventually we had we achieved it and and we offered a blog about what life is like on the other side of DataDog in terms of uh the cost savings and the performance improvements that we’ve received. And we we published that blog. I would encourage anybody who’s considering uh evaluating ClickHouse as an alternative to read. It’s deeply technical um and provides a bit of a framework on how to go through that process. But um I see DataDog being a very meaningful supplier. You you saw their recent quarterly results and some of the big wins that they were able to secure at some leading AI companies. Um and so I I do think that they’re going to survive this and and and come out on the other side successfully. Um but again, like like many companies, I don’t think they’re going to trade at the same multiple that they have in the past. Are you feeling an impact of AI on your hiring plans both on the engineering side where we’re seeing a lot of automation and on the go-to-market side where I I feel like there’s early energy around automation and it seems to have slowed down at least at this moment in history in 2026. You know, yes and no. No in the sense that we’re 500 employees today. We’re going to end this year closer to 1,000. So we’re going to double the size of the company in 12 months. And that headcount plan has not come down as a result of AI. Uh if anything it it may accelerate because our engineers are going to be 10 times more productive by embracing AI applications to make themselves more successful and their peers more successful and to develop code at at a much faster pace. There are functions in the company that we are automating that we didn’t even have beforehand. Like the concept of an SDR. When I started the company I said we’re never going to have SDRs or we’re never going to have CSMs because I think it’s clumsy and I think they’re functions that mask other issues around product quality and sales efficacy. Um and I always think about the customer’s experience and what it’s like to be handed off from a a junior inside sales rep to a quota-carrying rep to a renewals manager that’s described as a customer success manager, etc. But I think we can deploy AI agents that can serve a lot of those functions around lead qualification, demand generation, um enhancing the customer experience around giving them the information they need at that point in time, contextual help in the product, AI assistants, etc. The The ability to query your data using a natural language interface um through just some text prompt to say, “Hey, build me a stacked bar chart that shows revenue growth over the last 12 months and break it down by region.” I don’t need to go to a data analyst anymore and have that individual create that chart. I can just ask LibreChat, which is an open-source chat interface that we acquired last year and integrating with InfluxDB’s model and our MCP server into a ClickHouse cloud service. And Lucas, I would highly encourage you to check this out for your own implementation of ClickHouse where you can query your production data in real time and do ad hoc analysis. You can do situational awareness planning. You can do scenario planning. Like build me three different possible outcomes. Um uh optimistic, pessimistic, realistic. And it will go and query your data and and the insights you’re able to get from it are are staggering. So that’s kind of how we are using AI internally, uh how we’re building it into the product itself, which we’ve been thinking about basically before ChatGPT. Um but it’s not having a negative impact to headcount to answer your question. How is it affecting your engineering roadmaps? I mean I feel like infrastructure is the place where AI engineering has had the slowest adoption. Do you allow, for example, AIs to automatically submit code into your your codebase? There is typically a human review of code that goes into the codebase. You You have to remember, we’re building a database. We’re not like we’re building a mobile app over the weekend, you know, a dating app. Like move fast and break things doesn’t apply to a database. Like if we don’t have a highly reliable, endurable, and secure service, then we don’t have a company because our customers won’t trust us. And so it’s a little bit different, I think, when you’re developing this type of mission-critical application to just blindly commit code to it that an agent developed without some sort of human review um would be how I’d answer that question. But the the amount of contributions and commitments to the codebase that are AI-assisted is significant and is growing every day. And so I would anticipate if that number’s 50% today, within 6 months it’ll be 80%. Hm. All right. So I have another

[laughter] Here’s a Here’s a question I’ve I’ve been wanting to ask you. Um you’ve been having what looks like an incredibly charmed run. But as I’ve gotten closer to CEOs that that look like they’re on this hockey stick run, I’ve always found that there are these tough things internally, these kind of oh moments. I was wanting to hear from you what’s been the the biggest challenge or the the scariest moment in the life cycle of of ClickHouse so far. Well, it hasn’t been a straight line. I’ll start there. I may be from the outside looking in it has been. looks like a straight line from the outside, Aaron. I got to tell you. Um look, Yandex is commonly referred to as the Google of Russia. Which back in 2021 was not a bad moniker. It’s not a great one today. A condition of forming the company was that I had no ties to Russia. I couldn’t have any I didn’t have this foresight that Russia was going to invade Ukraine. But I knew enough they were going to be developing a cloud service and I couldn’t have engineers coding on a Russian network. I couldn’t have anybody in Russia. So I convinced Alexey and the team to move from Moscow to Amsterdam. That was not an easy achievement um to convince these engineers to pick up their lives and relocate to a new country. Then Russia invades Ukraine a month later. Like they moved in January of 2022. Russia invaded Ukraine February 24th of of of 2022. Six weeks later after they moved. So then we had this association and this relationship with Yandex that we needed to untangle. That was not an easy process. Let’s start there. So we celebrate the heritage of the software and the genius of Alexey and Sasha and Nikolai and Nikita and Ksenia and these incredible engineers that are still in our company. But we also sympathize with the fact that they basically had to leave their home country and relocate and have not gone back. So culturally that was very challenging. I then put $300 with Silicon Valley Bank and got a call that Silicon Valley Bank was going to default and I was going to lose all my money. And I said, “Oh shit.” Well, I’m pretty sure 200 million of that is off their balance sheet and being custodied at US Bank. I’m not entirely sure, but I’m pretty sure cuz Silicon Valley Bank was still the asset manager. But nobody could give me a clear answer on that Thursday. But I had a high degree of confidence that that 200 of the 300 million was not at risk. But there was a hundred million that was on their balance sheet. And I had this SWAT team, myself and my head of finance and my general counsel, trying to figure out like the rest of the industry, what the should we do? Like should we just wait it out and hope that they don’t default and we don’t lose a hundred million dollars or do we contribute to a bank run and wire the money anywhere, somewhere. Fortunately, we had banking relationships set up in Europe. Like many startups didn’t. They didn’t have a path out. So they were founders were wiring themselves money, which is not a great look. And so I I made the decision mid-morning on that Thursday. I called my head of Europe, Arno, and I said, “Wake up the bankers at uh at our bank in in the Netherlands cuz there’s a hundred million dollar wire that’s coming through in the next 30 minutes and it needs to clear cuz I think there’s going to be a massive run and the system’s going to freeze.” And so we got the hundred million out. I called a friend of mine at Silicon Valley Bank and I said, “Man, I’m sorry. Like I got to do what’s best for my company and my employees and our customers and I can’t afford to lose this money.” And so we wired a hundred million. It cleared. 30 minutes later SVB’s banking system went down. And people could not get their money out. Now, the government ended up backstopping SVB as we all know on Sunday, but there was 72 hours where nobody could tell me whether or not the remaining 200 million was safe. So that was a little bit of a a tricky time. As you know, being a long time ClickHouse Cloud customer, these services have issues. Like, you know, we haven’t had anything catastrophic, but around reliability and durability like we’ve had to turn inward and say, are we doing everything that we can to make this the most resilient service in the market? That has not been a straight line. I love where we are today in our security posture and our reliability and durability, but it’s not like you just wake up and have the most secure, durable database service. It requires a lot of sleepless nights and a lot of engineering work to make sure that you’re protecting your customers’ data and making the service available, especially for Weights & Biases you’re building a customer-facing application on it. Like you’re betting your company on our service. There’s a lot of responsibility that goes with that. I’ve been very fortunate with the team that we’ve assembled. For the most part, my leadership team is intact from when I formed it when we established the company and we added to it. So that’s been a total joy. Yeah, those would be the, you know, some of the trickier things that we’ve had to navigate in terms of company building, I would say. And then, you know, we’ve, you know, like Weights & Biases, we’ve got great investors. We’ve added to the to the roster. So so well capitalized. But yeah, I don’t think our company has really been tested in terms of how strong our resolve is and it’s going to come cuz the competition is waking up to the threat that we pose and they’re not just going to lie down and accept that ClickHouse is going to capture this market. These are very well organized, very well capitalized companies with incredible leadership. Yeah. Like who do you who do you see as as waking up and coming after you? It depends on the use case. So let’s talk about data warehousing and, you know, who are the big suppliers in that market? Snowflake is obvious. Google BigQuery is a great product with a big customer base that we would love to eat into. Amazon Redshift. Increasingly, while we partner very well with Databricks, we do see some competitive overlap. Those four. Let’s start there. With observability, I mentioned Datadog where there is some competitive overlap. I mentioned Grafana, Elastic. Again, these are great companies with with a very strong customer base with a lot of capital and with very capable teams. I feel like in terms of real-time analytics, which is what ClickHouse was originally designed for, we are the market leader in that category and I think, you know, whether it’s Vercel or Lovable or Sierra or Decagon or Ramp or Vantage or Klaviyo or Attentive, you name it. These are like customer-facing or Weights & Biases or Lane Chain. These are customer-facing B2B SaaS applications that I don’t think there’s a strong alternative to ClickHouse for those use cases. All right, let’s have a good spot to end. Thanks. Thanks, Aaron. Fun interview. All right. That was good. Awesome. Thanks so much for listening to this episode of Gradient Dissent. Please stay tuned for future episodes.