In this theCUBE + NYSE Wired: Mixture of Experts segment from the New York Stock Exchange, theCUBE’s John Furrier sits down with Raj Verma, CEO of SingleStore, to unpack how the intersection of technology and finance is shaping enterprise strategy. Verma shares why SingleStore is “on course” for the public markets, reflects on brand-building through the company’s partnership with golf Hall of Famer Padraig Harrington and connects that ethos to how SingleStore helps organizations fix struggling data “swings.” The discussion zeroes in on what’s next as Wall Street watches the AI infrastructure buildout: after chips and systems, the software and data layers set the pace for value creation.
Verma outlines why enterprises must modernize “brown” data estates into “green” ones to safely bring corporate context, governance and compliance into LLM workflows via RAG – and why commoditized data-at-rest puts the advantage at the query layer that unifies data in motion with data at rest. He predicts agentic AI will gain reasoning capabilities in roughly 18 months, cites industry indicators like Google reporting ~25% of its software now built by AI and argues that high switching costs will give way to disruption as buyers reassess legacy vendors. The conversation closes with concrete momentum: ~33% YoY growth, ARR in the ~$135M range, gross dollar retention ~98%, cloud NDR ~130, ~50% of business now in the cloud, landing ~3 new customers per day, a path to cash-flow breakeven in the next two quarters and a teaser for AI-related announcements in the next two months. Listeners will find notable stats, real-world use cases and forward-looking views on how databases power reliable AI at enterprise scale.
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
theCUBE + NYSE Wired: Mixture of Experts Series. If you don’t think you received an email check your
spam folder.
Sign in to theCUBE + NYSE Wired: Mixture of Experts Series.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open this link to automatically sign into the site.
Register For theCUBE + NYSE Wired: Mixture of Experts Series
Please fill out the information below. You will recieve an email with a verification link confirming your registration. Click the link to automatically sign into the site.
You’re almost there!
We just sent you a verification email. Please click the verification button in the email. Once your email address is verified, you will have full access to all event content for theCUBE + NYSE Wired: Mixture of Experts Series.
I want my badge and interests to be visible to all attendees.
Checking this box will display your presense on the attendees list, view your profile and allow other attendees to contact you via 1-1 chat. Read the Privacy Policy. At any time, you can choose to disable this preference.
Select your Interests!
add
Upload your photo
Uploading..
OR
Connect via Twitter
Connect via Linkedin
EDIT PASSWORD
Share
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
theCUBE + NYSE Wired: Mixture of Experts Series. If you don’t think you received an email check your
spam folder.
Sign in to theCUBE + NYSE Wired: Mixture of Experts Series.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open this link to automatically sign into the site.
Sign in to gain access to theCUBE + NYSE Wired: Mixture of Experts Series
Please sign in with LinkedIn to continue to theCUBE + NYSE Wired: Mixture of Experts Series. Signing in with LinkedIn ensures a professional environment.
Are you sure you want to remove access rights for this user?
Details
Manage Access
email address
Community Invitation
Sam Awrabi, Banyan Ventures
In this theCUBE + NYSE Wired: Mixture of Experts segment from the New York Stock Exchange, theCUBE’s John Furrier sits down with Raj Verma, CEO of SingleStore, to unpack how the intersection of technology and finance is shaping enterprise strategy. Verma shares why SingleStore is “on course” for the public markets, reflects on brand-building through the company’s partnership with golf Hall of Famer Padraig Harrington and connects that ethos to how SingleStore helps organizations fix struggling data “swings.” The discussion zeroes in on what’s next as Wall Street watches the AI infrastructure buildout: after chips and systems, the software and data layers set the pace for value creation.
Verma outlines why enterprises must modernize “brown” data estates into “green” ones to safely bring corporate context, governance and compliance into LLM workflows via RAG – and why commoditized data-at-rest puts the advantage at the query layer that unifies data in motion with data at rest. He predicts agentic AI will gain reasoning capabilities in roughly 18 months, cites industry indicators like Google reporting ~25% of its software now built by AI and argues that high switching costs will give way to disruption as buyers reassess legacy vendors. The conversation closes with concrete momentum: ~33% YoY growth, ARR in the ~$135M range, gross dollar retention ~98%, cloud NDR ~130, ~50% of business now in the cloud, landing ~3 new customers per day, a path to cash-flow breakeven in the next two quarters and a teaser for AI-related announcements in the next two months. Listeners will find notable stats, real-world use cases and forward-looking views on how databases power reliable AI at enterprise scale.
In this interview from theCUBE + NYSE Wired: Mixture of Experts series, Sam Awrabi, founder and managing director at Banyan Ventures, joins theCUBE's John Furrier to discuss how early-stage AI investing is being reshaped by insatiable enterprise demand and a new generation of capital-efficient startups. Awrabi explains why his firm targets pre-seed and seed AI-native companies across both infrastructure and vertical applications, with "AI-native" defined as building with state-of-the-art capabilities from day zero. He highlights watt-per-dollar efficiency as ...Read more
exploreKeep Exploring
What is the speaker's background, why did they start Banyan, and what does Banyan invest in (including how it defines "AI-native")?add
Why are pre-seed valuations and round sizes for chip and infrastructure startups so large, and what investment thesis explains your focus on those companies?add
Why is there such a large market opportunity for data center infrastructure, and how can companies like Argentum capitalize on it?add
How can a vertically focused AI product expand from a narrow use case into a horizontal platform or system of record?add
How do you identify which early-stage technologies or startups to invest in—i.e., how do you predict what will be hot in three to five years?add
>> Welcome back. I'm John Furrier, host of theCUBE. We are here in theCUBE's NYSC Studios, of course. We have a studio in Palo Alto connecting Silicon Valley and Wall Street Technology, is the market that's our NYSC-wired mixture of experts, programs, theCUBE original where we expert the leaders who are doing good work in the field of AI. We got Sam Awrabi, he's the founder and managing director of Banyan Ventures, Solo GP, and a bunch of hot deals, cloud, Neocloud, GPU clouds, vertical, AI infrastructure. Great to have you on. Good to see you. Thanks for coming on.
Sam Awrabi
>> Thanks, John. Great to be here.
John Furrier
>> So you're investing in two hot areas, AI infrastructure and vertical blank apps, infrastructure agents are going to do well there. Talk about what you're investing in as a solo GP. What do you do and then what do you see? Start with what you do.
Sam Awrabi
>> Yeah, absolutely. So my background, starting in 2018, I was the first technical sales hire in what's now a $4 billion AI infrastructure category, which is experiment model management. So spent four or five years meeting two or three data science teams a day. That was my platform, those relationships. I took those relationships and launched Banyan to... Instead of providing guidance or consulting work or full-time work, to provide them capital plus all those other things. Banyan, my firm focuses on pre-seed and seed AI-native infrastructure and AI-native vertical application software. And I define AI-native as building with state-of-the-art AI capabilities from day zero as you launch your company. So on the infrastructure side, there's a lot of research out of PhD programs that goes into infrastructure. So that's AI-native there. And on the vertical side, it's targeting unique data sets, unique workflows and building businesses around that.
John Furrier
>> It's interesting, if someone's watching, "Oh, you're investing in AI infrastructure, pre-seed, what the hell does that even mean?" Because people are getting billions of dollars to build out all these factories. So explain the nuance of the entrepreneurial opportunities in AI infrastructure. It's not just what people think. Billions building nuclear reactors, there's a lot more going on. Take me through the mindset of the market that you're investing into, because there are pre-seed deals that can grow fast certainly, but it's not, you're putting in a billion dollars of pre-seed. So some of these pre-seed deals we're seeding for chips and whatnot are like a billion dollars.
Sam Awrabi
>> Right. So valuations are all over the place, round sizes. But in infrastructure, the first principle is watt per dollar efficiency. AI takes a lot of energy, and the last 20 or 30 years have been dominated by general purpose compute, compute that could be used for every which use case. And now we're moving into a world of specialization where across this entire infrastructure stack, there's highly specialized infrastructure being built that has much better efficiency, much better watt per dollar efficiency. So companies, pre-seed that I'm investing in typically have a lifetime of research behind them, and PhD programs, they're very technical founders. They're four or five years ahead of the current market status quo in their respective fields. So examples are biological computing, radio frequency chips, inference chips, and a lot of others. One of them Argentum AI, a decentralized GPU cloud. I led the pre-seed around there. So yeah.
John Furrier
>> Also on the chip side we had Mitesh on.
Sam Awrabi
>> Yep.
John Furrier
>> You're also an investor in his company?
Sam Awrabi
>> Yeah.
John Furrier
>> Positron is... Ah, he's got the best hair in the Cube alumni-
Sam Awrabi
>> He does have good hair.
John Furrier
>> He's got good hair.
Sam Awrabi
>> But yeah, I mean what's interesting about Positron is, for example, when I'm saying this ultra-efficiency watt per dollar angle, their chip has around two to three X more memory than other generations of Nvidia hardware on the market now. So that brings the chip inward on the data center, less power, less energy, less real estate, and actually improves the inference speed because when you're on GPT, you don't want to be sitting there waiting for GPT to load. And yeah, I mean Jump Trading co-led the series B, a leading hedge fund. They're also a really happy customer. We also had other great investors, like Arm in that round, Qatar, so it's a really exciting company for sure.
John Furrier
>> And Argentum, we've had Andrew on, the founder of that company, very innovative company, really growing super fast. I mean that's opportunity recognition right there.
Sam Awrabi
>> Yeah. So with Argentum, I met him, led his pre-seed round, we're the largest shareholder in the company, Kraken followed on. I believe you are also an investor, which is amazing.
John Furrier
>> Well, small taste, but that to me signifies the movement. You're starting to see two forms of infrastructure, pure hardware commodity, not commodity, but differentiated by just watt per power, token per watt, whatever it is. And then you got the specialized, kind of like SLA driven, I want to win the corporate enterprise or provide unique value add to that. Is there an opinion on your side of... I mean they're not mutually exclusive, they serve two things, but how do you view that market? It seems, I won't say bifurcated, but it seems to be two real clear value propositions. Get me power and compute and horsepower, token, watts per second. SLA that up. Then you've got the corporate SLA around my workloads, agents, the SLA on the Argentum side, the Neo cloud side is like, "Hey, okay, this software works. I'll give you what you want." Versus my workload needs this security posture, I have agents flying around. They're two different scale.
Sam Awrabi
>> They are. So if we take a step back, there's around $6 trillion of infrastructure that still needs to be built and come online. And that's an astronomical amount of land, power, humans that need to go build data centers. And so the market size itself is one of the largest build outs ever in human history. And so companies like Argentum are coming in and saying, "Hey, we can actually move a lot faster at finding land, finding power, having a customized SLA that fits that enterprise's needs." And we have a very clean balance sheet. We're not overly debt ridden. We aren't in a lot of scenarios yet as an organization that are not advantageous to us as a potential infrastructure builder. And we can move extremely quickly. And that's what Andrew's doing now. And that flexibility to really move fast and be creative sparks very large potential.
John Furrier
>> The demand curve, I mean just go to the demand, the demand side's off the charts. Talk about the verticalization, because AI's about domain expertise, domain data, specific workloads. Agents are super hot right now with all the action, physical AI, which we've been covering here is hot, you got robots, it's an edge, Nvidia super computing vibe, powering all aspects of this vertical. You mentioned the matrix. We use that all the time. I mean you could go to an agent saying, "Give me how to fly a helicopter," famous scene where it uploads the capabilities. Judo, those scenes in the matrix, that's happening now with AI. Be my professor, be my ghostwriter, be my coder, be my marketing lead, be a lawyer. I mean all this is happening. What do you see there as the opportunities and what you're investing in the vertical nature?
Sam Awrabi
>> Right. So my view on the world with vertical AI is what starts as vertical becomes horizontal. And so what I'm looking for are highly specialized vertical application use cases that access hard to reach data and hard to reach workflows. And once they're integrated into that data and that workflow, that's a very meaningful workflow, whether it's the New York Stock Exchange on running compliance or there's all these different examples. Once you're in that data, once you're in that workflow, you can then become horizontal and become the system of record.
John Furrier
>> Explain that again. Rewind, because I think that was a key point. Explain that concept again, that makes it work.
Sam Awrabi
>> So for example, one of my investments is a company in Italy. I was the only US-based venture investor called Lexroom, and they focused on only so far the Italian legal market. But language is a moat. So for general purpose LLMs, even Harvey AI, these other legal players, they haven't yet gotten to the level of accuracy in the Italian market. Italy actually has the most lawyers out of any country in Europe. So just that one market is actually pretty large. So they go in very vertically integrated on the language and on the workflows of lawyers and they're having a lot of success. Raised the series A, more fundraising to come, and now they're expanding into other countries in Europe with ultra language-
John Furrier
>> So what was once a side hustle or a lifestyle business or cashflow funded business, you now can come in on a very narrow entry, lock something down and sequence horizontally.
Sam Awrabi
>> Yeah, exactly. Or like other-
John Furrier
>> What are some of other examples that you see?
Sam Awrabi
>> Yeah, so in my portfolio there's Examine, they're doing ultra-specific AI for commercial real estate managers, like very large commercial real estate managers. Another round, I was an investor in ATG Group, Garry Tan from YC led the round, the co-founder of Jane Street, CEO of Nebius. They're replacing financial advisors. But think about all of your estate, all of your data from your financial life going into an ultra-specialized AI. You don't need a financial advisor giving them a two or 3% fee. AI can replace a lot of that decision making and organization of high net worths. And that's like another example. So these are very vertically focused companies going in on mission-specific workflows and then they quickly expand into operating capital.
John Furrier
>> And so there's two paths of liquidity, M&A, so Harvey would buy your portfolio company-
Sam Awrabi
>> That'd be a great-
John Furrier
>> Or they sequence and own Europe in every language sovereign, it's a kind of sovereignty AI.
Sam Awrabi
>> So my first ever investment was acquired this week by Brookfield Radiant and it's a hundred billion dollars infrastructure build out. I was one of the first five investors in Ori as he pivoted into the GPU space. And that was the initial thesis, is that the UK needs a sovereign GPU cloud. So things went differently than maybe that initial thesis because he closed great deals in the US, like Together AI, SF Compute, et cetera, et cetera. But for Brookfield Radiant, they have land, they have power, they want to build a very vertically integrated GPU stack. And the Ori team was a great partner to them to help accelerate that. It just made sense to partner together and acquire the company.
John Furrier
>> What's your outlook now as you look at the market? We can see the role of Nvidia has become really strategic in the AI infrastructure space. A lot of bankrolling investments. I think some people look at it as kind of inside dealing. I really don't. The accounting scandal, or the accounting issue of revenue recognition in the early dotcom bubble days, that was an accounting thing that was solved by SOP 97-2. I remember it vividly. So revenue recognition was born. I just think there's no real comparable to what Nvidia's doing. I think you can't really compare them to what no one's ever done this level of architecture from a venture standpoint, growth, the supply chain. I mean Intel had their run, but it was a completely different paradigm. So the fact that Nvidia's making so much money, as is AWS and others, why don't they play a role as a capital provider, or they're motivated. The more AI adoption, the more applications, it's a virtuous cycle. I have no problem with Nvidia. I mean they're making the market.
Sam Awrabi
>> Right. So I mean, taking a step back, Nvidia's worth more than all of big pharma combined. And everyone in this room today, if we went and did a poll, they'd all say, if you asked, "Hey, how many GPUs do you want for your AIs now and in the future?" No one would say zero, most likely. So the GPU build out again is so massive that NVIDIA is the predominant market player, 90, 95% market share. And if you look at their stock, even with the most recent earnings announcements, the margins are sustaining, the revenue is continuing to beat all expectations, I think 12 quarters in a row, yet the stock is actually going a bit down on sells to market value. So I think-
John Furrier
>> There's some profit taking in there, but I still, in the long game that they have, I just don't see that fading anytime soon.
Sam Awrabi
>> No, but I do think there's room for specialization, like what we talked about with Positron, and this is happening across the infrastructure stack. Other parts, there's different use cases inside of inference that need specialization and those are their own sub, potentially billion dollar margins-
John Furrier
>> Sam, I want to get your thoughts on this because I've been trying to compare this bubble or this transition. The threshold has been crossed in AI, in my opinion. So the genie's out of the bottle. People compare it to the dotcom bubble and some even reference to the telecom bubble that popped, over building of fiber, Quest sold to, I just had Global AI in and Lumen, Lumen bought Quest, the fiber. Those demand curves of the internet and telecom weren't there. So the over-build or the hype of getting all that data center and connectivity for the internet, the web in 1999 to 2001, that was... It was accelerating, I mean, Mary Meeker used to keep a stat on online population. So it was growing. So that was embryonic with evolution. Okay, telecom bubble, no demand for services, not a lot of bandwidth. Web pages had barely any graphics on them. Now you got the infrastructure, the cloud, telecom, multimodal content, the bandwidth needs are there. The demand curve for AI apps are high. So I don't see the demand curve dropping anytime soon. So I think all the fallacies of what could happen, in my opinion, are a little bit need to be weighed. I don't see a bubble popping. I think the music might slow down in the musical chairs analogy, but that's just the nature of the game. But there'll still be a need for more and more data centers.
Sam Awrabi
>> So I think in past markets there was a bit of a build it and they will come mentality. I wouldn't say wishful thinking, but hey, let's build ahead of the demand and the demand will come. Whereas in AI, if you go into any enterprise organization, top down in the boardroom, they're saying, pounding the table, "Get AI in every workflow now." So there's such a insatiable demand across all parts of our economy to adopt and apply AI. And so there really is no... That's the difference now. And so even in my portfolio, from pre-seed and seed alone, there's 600 million in revenue bookings and that's across like 23 companies. So the revenue per employee ratio is so much higher than the software companies of the past. The revenue growth, there's multiple companies that have done zero to a hundred million in less than 24 months in the portfolio. And most of those companies are not even a hundred people yet. And so we're re-architecting what the new 1% of outlier startups look like in real time due to AI and the demand underneath it. So I don't think we're in a bubble, but there'll be a lot of big lose-
John Furrier
>> Yeah, your observation... Well, there'll be losers, people who don't build a good product.
Sam Awrabi
>> Yeah, well there's a lot of funding, a lot of horses in the race, and only one horse can win, and most markets are winner-take-all. That's how-
John Furrier
>> We had Matt, a Cube alumni who's got the mill venture, former Ness founder, Apple with Steve Jobs before, and I lived in Palo Alto for 25 years, no other jobs, family, and know a lot of Apple people. The Apple is the poster child for what you just said. The companies that will die from AI are the ones that just don't evolve. iPod, the hottest selling product revenue wise for Apple was in 2005, 2006, 2007. The iPod generated more revenue than anything else in Apple. So what happens in 2007? The iPhone comes out. Conventional wisdom will say, "Wait a minute, we're making so much money with the iPod-"
Sam Awrabi
>> Why go to this new market?
John Furrier
>> They collapse everything in, it's a phone, it's a computer, it's an iPod, it's got an app store platform built in. They just created a better product, no key, or rim, or sit in their BlackBerry. They had all the piece parts. They did not do that. They died. Apple won. So to me, whoever doesn't align with the interfaces voice, going to be some agents. But there are moats.
Sam Awrabi
>> Right.
John Furrier
>> There are moats for a company saying, "We can actually pivot and eat our own."
Sam Awrabi
>> Right. And I think part of my job is, it's not what's hot today, it's, I'm at inception stage, very early in these companies' life cycles and markets saying what will be hot in three to five years? And so for example, one of my investments, Akash Networks, they just announced the first ever diamond cooled server kind of cloud ever in history. So they're literally putting synthetic diamond in a server cooling at 1.5 x compared to liquid or air cooling. And so as we potentially move data centers into space, a new cooling frontier using synthetic diamonds actually is a very useful technology.
John Furrier
>> Who would've thought about that?
Sam Awrabi
>> That's just a great example.
John Furrier
>> This is a great example of the white space opportunities that can go horizontal. You don't know. All right, I brought the Apple thing up that I want to get your thoughts on this because that's the paradigm. The companies that don't succeed are going to fail. Okay, so that means that there's going to be that 1%. So okay, IBM's pivoting. So if you can do that transition like Apple did in the big company, it's productivity, not how many employees you have. So you're seeing that.
Sam Awrabi
>> Yeah, it's revenue per employee.
John Furrier
>> Is a great metric.
Sam Awrabi
>> It's your net margins after all of your-
John Furrier
>> So that's for the big companies.
Sam Awrabi
>> Right.
John Furrier
>> Innovate or die, new metrics, all the other metrics are changing. But for startups, you mentioned this new 1% trend, or unicorns or decacorns, whatever you want to call them, that there's going to be a class of startups that will look like the one percenters, but now that's in the power law as the 20 percenters. Is that what I'm getting at? So you're starting to see a redistribution of what we used as the holy grail one percenters.
Sam Awrabi
>> It's all changing.
John Furrier
>> Explain this concept.
Sam Awrabi
>> So Andreessen Horowitz in the past year, I'd say, came out with a data set that they procured, and a 16Z, arguably the best VC in history. They're saying, "Hey, for consumer, the new 1% is zero to 12 million in revenue the first 12 months of you launching your product, and for infrastructure B2B, it's zero to 6 million the first 12 months of you launching your product."
And so in the formal world, even that I used to be in, in MLOps or AI infrastructure, if you could launch your product and get to 1 million in revenue in the first year, you were at the top 1%. If you could get to 3 million, that'd be like, "Oh my gosh, that's unheard of. You're going straight to your series A."
And so just the revenue trajectories are changing and what the top 1% looks like. And then if you go to the really 0.01%, the true outliers, like OpenAI or Anthropic, I mean their revenue trajectory from launch, and it's public knowledge, you could look it up, the world has never seen anything like that from a company perspective.
John Furrier
>> And as an investor, you're spotting it out. Well, great to have you on and thanks for coming on theCUBE, great to see you. Love the entrepreneurial opportunity recognition. Investors are also changing what deals look like, what success looks like. All the metrics are changing. At the end of the day, it's the same old game. See an opportunity, go after it as fast as possible. Capture it, lock it down, grow it, make it durable. Sam, thanks for coming in.
Sam Awrabi
>> Thanks so much, John.
John Furrier
>> Banyan Ventures doing a lot of great work. I'm John Furrier with theCUBE. Thanks for watching.
John Furrier