Lambda: From Convincing Buyers to Use GPUs to an 11-Figure Deal | with Robert Brooks, CCO

0:00If anyone can build anything, who the fuck is going to sell it? We are. We are going to sell it.

0:05Sophie Buonassisi: Robert Brooks, Chief Commercial Officer at Lambda. There's nothing really neo about Lambda. Lambda's been around since 2012. There's 5 ,000 net new companies that sign up for our platform on a monthly basis. Let's say you've got 10 grand. You can spread it any kind of ratio you want across SpaceX, Anthropik, or OpenAI. Where are you cutting the money? Wow. Are those secretly AI glasses? Very stylish ones, right? You're observing this thing in real time. Yeah. Yeah.

0:43Sophie Buonassisi: Robert is the chief commercial officer at Lambda, one of the companies providing compute behind the AI boom. He is incredibly close to all the AI infrastructure that is being committed right now has that really, really interesting close perspective on it. He's also a GTM Fund LP. We're incredibly grateful for him and happy to be sitting here on stage with you. Excited to dive in. Thank you so much, Sophie. Before we get started, I wanted to personally thank you and thank you so much to Paul. Thank you so much to Max. There's been two ways that GTM Fund has really affected me in such a positive way.

1:17One, I made my number two hire in my organization from this community, just from printing it out on Slack. That type of like value is insane. This is my number two guy, right? Former head of sales operations at Slack. Now my head of go-to-market operations at Lambda. Incredible value add. Number two, got to do a podcast with you folks a couple of years ago. And that helped validate so many ideas that I had stuck in my head and get it out there in the world. And it was really fun to just have those opportunities. So thank you guys for that.

1:50Sophie Buonassisi: I appreciate you sharing that. and it is our absolute honor and pleasure. Now, for anyone unfamiliar with Lambda, give us, in a couple sentences, a bit of an overview of Lambda and your role there. Yeah, so we all wear white glasses and we all wear white seer sucker suits. Very stylish. After labor. And the shoes are from Portugal, so you just gotta, you know. Anyway, so we're a NeoCloud, as it's known. I don't necessarily love that framing, but it is what it is. You kind of have to lean into it these days. There's nothing really Neo about Lambda. Lambda's been around since 2012. In fact, I joined before AI was a thing that was commonly said.

2:27It was very cringy in 2018 to say AI. It was machine learning. It was deep learning. So today we build superintelligence, sorry, supercomputers for superintelligence, big, massive data centers, hundreds of thousands of GPUs. We started building gamer workstations with a little bit of software on top. And now we're here today managing fleets, you know, running some of the biggest agent swarms in the world. So it's been a lot of fun.

2:51Sophie Buonassisi: That is fantastic. And so who's buying compute now that maybe wasn't, say, a year ago? And what are they asking for? I think the dude working at the hotel is buying compute. Like, everybody's buying compute, right? And so, like, you know, honestly, this market in 2018, I had to convince our buyers to use GPUs, which is sort of an insane statement today. Today, there's a lot of inbound demand. And my team is much more less on the hunter side, much more selecting the right customer to grow with and scale. Right. And so when you look at a business model of any company like in this portfolio or in general, it's adopting AI and needing to use GPUs, not just relying on an API.

3:32It's certainly about your business model and how you're going to scale that business model and us having a limited capacity of something and wanting to scale that with you because we think you're going to be big. So it's almost like this venture capital take on like selecting a customer. Now, it's a privileged position to be in. And certainly it'll change in 2014. In 2024, we had a glut of GPUs and couldn't figure out a way to give them away. So there will be busts and booms here. But ultimately, our buyers consist of obviously inference providers that are out there that do really, really exceptional work.

4:04I'm sure some of you in the room, maybe your engineering teams have used our platform in some sense. The big frontier AI labs, the hyperscalers, hedge funds, this is predominantly where I spend my time. And the actual, like, the persona is what's really changed for us. So we used to sell to machine learning engineers who, as you know, meta hires them for a billion dollars these days. These guys have a lot of economic pull. They can just make a decision and the business moves towards it. That taught me something really early on in my go-to-market career in terms of selecting a user profile to build towards.

4:41They have economic power, the machine learning engineer. And so they can make, you know, maybe a VP said no, but if the machine learning engineer said yes, the deal happened. So that's still generally true today, but today it's obviously in a much bigger stage. Deals are in the billions of dollars. I've done an 11-figure contract in the last three months. That's a lot of zeros. my hairline is now further back my waistline is now further out um that's a true story by the way but like uh so we're now at the sort of like founder level when we're you know selling to ai natives and we're at the sort of like head of data center construction and head of compute these are titles that didn't exist 10 years ago uh well data center did but head of compute didn't really do at major companies.

5:31So the title's really shifted. The persona's shifted. The buying cost has really shifted.

5:35Sophie Buonassisi: A quick pause. If you run outbound sales, this is big news. Most reps are bouncing between five different tools just to send one email. But there's a better way. Nooks is the AI agent workspace built for intelligent outbound, where AI agents and your reps work together in one place. Nooks' AI sequencing handles the busy work, finding prospects, writing outreach, knowing who to contact and when, so your team can focus on actual conversations. If you're evaluating your outbound stack or renewing legacy tools, check out nooks.ai. That's N-O-O-K-S dot A-I. So it's shifted a ton. You mentioned there can be some lumpiness to it.

6:13Sophie Buonassisi: You have purview to who's actually buying Compute long before typically, you know, many of us see it on the product side. What is what you're seeing telling you about where the demand is heading overall. Yeah. I mean, in terms of venture, I would say that we have the greatest deal flow in the history of AI. There's 5 ,000 new companies that sign up for our platform on a monthly basis. Extreme amounts of demand, right? The number one thing that we're seeing that we're really liking from startups, from true startups, right? Not just a startup that raises a $2 billion seed round, is the verticalization of AI agents or just like an AI platform in general.

6:53So drug discovery in a particular sort of like molecular approach to solving something very small, if they're applying a path towards that, you know, that's a really cool thing that we're seeing that's really starting to take off. The agents thing, like obviously all over the place, very, very big on that. Believe it or not, agents run on CPUs. So our business has changed as well. We're now deploying hundreds of thousands of CPUs. Didn't think we'd be doing that a year ago. So I think the quote from like, if you can predict the future today, you're insane. like you can predict next week, you know.

7:26Sophie Buonassisi: If you could predict next week, what would it be? Next week, I hope that it's not 95 degrees in San Francisco. That's what I predict, or that's what I hope. But honestly, I think next week we're going to find out a little bit more about Anthropix's IPO. Yeah, yeah, very true. I mean, with that in mind, I have a little bit of a curveball question for you. Here we go. Robert. Hypothetical question. Let's say you've got 10 grand. You can spread it any kind of ratio you want across SpaceX, Anthropic, or OpenAI. Where are you cutting the money? Wow. Well, I guess SpaceX is the only one that's public, right?

8:02So I'd have to find my way into an SPV stocked onto an SPV stocked onto an SPV to get into an Anthropic round, probably, right? All kidding aside, I mean, all three of them are just generationally amazing companies. and I have to be biased because I'm an unbiased platform. So it'd be a clean 33-33 split, 33-33-33 split between all three of them. Massive respect for all of them. I love the open-air approach towards consumer. I love the cloud approach towards enterprise and a specific software domain and the SpaceX element of like building data centers actually in space. However insane that sounds, like I almost like, you know, spit up when I heard that idea six months ago.

8:47I totally believe in that idea now. It's going to be generational. Space is big. It's going to be fun. And there's free energy.

8:55Sophie Buonassisi: That's a very investment forward approach. The 33.3. I have to be unbiased in that answer. Yeah, yeah. That was my guess of what you'd say. And probably a pragmatic one, how we would do too, but had to ask. It would be unthrobbing. Now we've got a room full of many go-to-market operators here. You hear yourself relieving the commercial side of Lambda. And what kind of advice are you giving your team? And are you seeing other go-to-market leaders finding value in on the go-to-market side, given everything going on with AI? Okay, so in general, I think that the people in this room and the skill sets that we have are going to be the most important thing in the next 10 years.

9:34Bar none. If anyone can build anything, who the fuck is going to sell it? Sorry. We are. We are going to sell it. that skill set that you have that you may not even know that you have, that you were born with, is one of the most important things that's going to completely take over this Silicon Valley, maybe a super intelligence valley. And I think the demand and the value of our skill sets are going to go parabolic. To sell, to build relationships, to build trust, to get someone to convince you, I'm sorry, to convince someone that you are not bullshitting them is a property that a zero to one bit will never have.

10:19We have it. It's in our DNA. The people in this room are some of the best people ever to do it. That's my advice.

10:26Sophie Buonassisi: That is fantastic advice and very promising for everyone in this room. Last question for you, Robert, what is your favorite personal agent use case right now? So my world is sort of shifted in so many domains. I don't get to verticalize in a particular area of focus. I have to be platform agnostic. So one time I'll meet with a customer in manufacturing, I'll meet with a customer in bio, then I'll meet with a customer in semis. So I'm still at the prompting stage of like just having to in real time, get up to speed on everything that these people are doing, basically while I'm in a meeting with them, especially if it's virtual.

11:03And so I think that's an incredible unlock. If you remember selling back in like before these times, if you didn't have a prep doc, you were screwed, right? You had to Google and then maybe you'd find the link and you would just, you would not show up in a meeting knowing like an expert. You can quickly turn around in about 45 seconds and sound like an expert with a prompt. That's a key unlock. And it's so simple. And I see my reps not doing it, but honestly, if it's right there at your fingertips, take it. So for me, that's like how I utilize agents the most. And they're not really agents. It's still at the prompting stage of that right now.

11:37Sophie Buonassisi: Are those secretly AI glasses? Very stylish one, right? You're observing this thing real bad. Yeah, I have had, I think that one thing I have had the pleasure of getting attuned to is having a robot. So we have, I use, I have$40 ,000 left over in my marketing budget a couple of years ago. I bought a robot. We have a machine learning team that programmed it and we got it to do things that we didn't want to do. I don't want to. How tall is this robot? Let's get a visual. It's about like 5 '2". Okay. Yeah. Maybe it's shorter than that. Maybe it's like 4 '10". But we've programmed this robot to like literally just like take the trash off our desk.

12:13And I can't like begin to explain how important it is when you have 50 people working in office, like a robot doing that for you. All of a sudden, like things are just more tidy. Like you, you know, you never really yell at anyone. No one has to do like a PSA that you should like wash your dishes. Like this robot was doing it. so I think like all of us are going to be able to experience that in the next coming years. And it's going to be a very, I think, cool thing to come into our everyday lives.

12:38Sophie Buonassisi: Very cool. We're, we are going to need a picture of the robot. While you're not taking the selfie, we need that picture of the robot sometime. You got it. Awesome. Robert, thank you so much. Thanks everybody.