#217 CEO & Co-Founder Codeium, Varun Mohan w/ Leigh Marie Braswell: Limitless

I was reflecting on our first dinner together. It was shortly after our Series B investment, and it was me, you, and Graham, your new sales leader at the time, was your new sales leader. And well, first we like changed restaurants 3 times cuz you're like, Juwan, not this restaurant. How about this restaurant? This restaurant? I'm like, geez, like I haven't even really met this guy yet. We go, we sit down, and immediately you and Graham are kind of like negotiating the number. Like, like, he hasn't even started yet.

He had started. He might have been there for like a couple weeks at this point, if that.

This was his first trip to the Bay Area to like spend time with the team, right?

That's right.

And you guys are like negotiating the, the goal for the year, and the number was ludicrous. Whatever the number was, it was, I think, like 15x growth for this year. And I remember, I think I texted Leigh Marie after and I'm like, this guy's nuts.

You did, you did text me that.

And I actually felt bad for Graham, 'cause, you know, the sales leader in me and I'm like, this guy's signing up for a path that leads to nowhere here. You were like, no, we're gonna, we're gonna do it. And then, you know, Graham, to his credit, was like, well, I'll sign up for it. Like, let's do it. And you've done it. Like, you've actually done it. Like, you're— and not only have you done it this year, you're like blowing that original number out. Like, this business is gonna grow 10+X this year. How crazy is that?

Yeah.

Okay. Is that, is that only crazy to me? Like, did you actually, did you think you were crazy then too and you knew it? Or does it also like, or is this just like, this is what it felt, this is what you thought it was gonna feel like?

So I think it was probably some amount of irrational optimism because, you know, you can kind of see the pipeline, right? You can kind of make some assumptions about if a bunch of things close, it'll actually matriculate. But then there's this entire process of actually building out a sales team at this point. By the time we sort of got Graham, we had closed a couple large sales deals without Graham already. So I think we projected out every rep that we'd be able to have would probably be able to do the same. Obviously, a founder or a series of people that were on the founding team selling the product is very, very different than a sales rep selling the, selling the product. So that was irrational on my part. But I don't know, maybe, maybe there is a little bit of, if you, if you try to goalpost yourself based on like what other companies do, which is the triple, triple, double, double. you're kind of going to get like a middling outcome anyways. So you might as well set the bar really, really high because you may actually exceed it, right? We had already heard of companies like Wiz that hit ridiculous numbers. So it wasn't unheard of. But in retrospect, I think a lot more things needed to be built than I thought were really necessary. And you kind of learn along the way. I'm sure like we're going to set a crazy number for next year.

You're already talking about it.

Yeah.

You're doing it. You're doing it again.

Yeah.

You're doing the exact same thing again.

Yeah, that's probably true. Yeah.

And I remember Leigh Marie, like we were, I think we were like, look, if they do half of what they wanna do, they're still gonna be on a trajectory that is in a class of its own.

Totally.

And we're like, all right, like that seems fair. You know, like, and we're still like this guy, like what I thought, and I'd be curious to hear what you think. I was like, either he is, 'cause every founder says, We're gonna go do this insane goal, right? Everybody wants to do that. And once you did it, I was like, holy shit, holy shit, they actually did it. And for me, it went from like, maybe he's a little nuts, to, oh, like, he's gonna say what he— he's gonna do what he says he wants to do. And I don't know, for me that was really refreshing. We don't always, we don't always get it backed up in the same way. And, um, maybe he's not so crazy.

I will say, I think all the best founders are at least a little crazy.

Yeah.

A little nuts.

You think so?

Um, it takes a little bit of a crazy person to, to be able to do this job. Um, but yeah, I'll say 2 things about, about the crazy number and then us, um, us hitting it and probably setting another crazy one next year. I guess one, I mean, I think it's just a testament to Bruin and the team's ability to just like— I mean, they're one of the more dynamic teams I've just ever met. And we can obviously, you know, talk more about this later, but they just have a history of being able to like see an opportunity and seize it very quickly and do things themselves that might be outside of their, you know, normal area of expertise. And I think the very best teams and founders, it's just very clear that you can see their slope and how they improve across multiple dimensions, whether it's setting up enterprise sales processes and doing things before there's even really any sales team in place. It's just figuring stuff out. And I think Varun and the team just kind of has a demonstrated history of doing that. And then 2, there's this interesting trend now. I want somebody to redo the triple, triple, double benchmark because of like what we are seeing now in AI. I think there's both a combination of there's this just sort of massive amount of interest from consumers, from enterprises, from everyone to see the full power of this technology. it's only been out for a little while. Like ChatGPT only came out 2 years ago.

Yeah.

So you've got like, it's very new. There's a ton of opportunity and you're just seeing teams now kind of like blow past the triple, triple, double and with fewer people because they are also able to use AI internally to sort of streamline processes. So I would— I think that rewrite it. Yeah, I think the very best founders can rewrite it.

And to be clear, when we say triple, triple, double, double, like this is the trope in venture that's like you go from your first million of ARR, you'll go to $4 million, from $4 million to $12 million, $12 to $24, $24 to $50. And that was like for a long time best of breed. But you're gonna do the whole triple triple double double in like a year and a half, not 4 years. And, um, yeah, maybe we do need to rewrite it. It's not like— it's funny because, Varun, not just you but your team always messages me like, are we the fastest? Are we the fastest? Like, is anybody growing faster than us? How quickly did Wiz grow? This, like, uh, this ambition permeates through the org. Do you feel that?

Yeah, I think we do have one of the more like competitive teams, I would say. Not even just on the sales side, but even on the engineering side. And maybe it's kind of like— and, and, uh, maybe let's, let's take the engineering team. I don't think the engineering team knows any better. It's not like they know that, hey, the limit is triple, right? I think maybe sales knows that at some point they heard that triple is a great number, but if you don't know the limit, then you shoot for the highest number you can. And anything kind of just feels reasonable at this point. I always think about it kind of interestingly. If anyone leaves our company at this point, their expectations on the next company are going to be ridiculously high. They're not going to be, hey, are we—

At least Cody, if you're saying. Yeah.

They're going to, they're going to at this point be like, oh, like we grew the business by 2x, it's going to be like, okay, that's not that great.

Yeah. Well, that's actually a bug, not a feature of people that come from super successful companies because they've seen what world-class looks like. And so therefore, they've raised their own expectation of how everything should feel. And when it doesn't feel that way, they just throw their hands in the air being like, well, that's not the experience that I had at my last company. You know, that's like, like the delta between your, your, your reality and the expectations. Just like people just quit, just leave, you know?

Yeah, I think, you know, one of the big things that I kind of took away with this experience is product market fit is you can kind of feel it when it's, when it's real. And like, probably everyone that leaves our company is never going to kind of fake feel that, hey, this company has product market fit, which is that we kind of limped through over the quarter, over the year. They're gonna know the difference, which is in and of itself like a very useful thing. Like, I see this constantly where you take these people that went from Google and they try and start a company or they work at a company and they kind of don't know that Google is the greatest business of all time, right? Like, I, I try to tell other people Google's like this, this downhill slope, right? And as long as you're not a ball of mud, you're gonna go down the slope really, really fast, right? You could become a VP as long as you can talk to people well. I'm, I'm simplifying. a little bit, right? I'm not trying to, trying to downplay all the VPs at Google. But at the same time, when these people leave and start companies, they don't know what the baseline is because whatever baseline they write, it's going to be way worse than Google. And so, they don't know what, what great looks like because they've only seen awesome. You know what I'm saying? So, because of that, they kind of just like flail around. They don't know how to even make something that even is 1% as good as Google.

But do you think The company that you were at before Codium, which is, I guess, your only job, right? The only non-internship job at, at Neuro, correct?

Yeah, that's right.

And like, would you consider that a successful company?

The story hasn't completely been written yet, but I would say that the company was supposed to commercialize much sooner than it has, right? It has still not commercialized the technology. Just to set some context, it's an autonomous goods delivery company. The mission of the company was to build an autonomous vehicle from the ground up that does things like DoorDash, right?

Yeah.

A little bit different than Waymo in that Waymo transports people. And I think that was actually maybe a really, really useful experience for me because it taught me, hey, what is it like to work at a hard tech company? Because I truly do believe Codium is a hard tech company, right? The things that we aspire to do are a little bit different than HR software, right? Like, I don't know what the future of our company is going to be. I don't know how to solve most of the problems at our company. Most of the technology for our company has yet to be created. The same is true in autonomous vehicles. But I think the biggest learning lesson for me from Nuro was how do you actually build a hard tech company that is actually making incremental progress along the way, is actually profitable along the way, revenue generating? And we weren't able to crack that at Nuro.

Yeah.

And I think what I was kind of searching for, and I've always been searching for this, is a company with a grand vision that solves hard technology problems, But actually can make progress along the way, right? I don't know how to build a business like Neuro.

Doesn't that make it so much more satisfying knowing the other side though? Like, for example, my first job was at an early cloud security startup, raised a bunch of money, never found product-market fit, did one big deal with Goldman for $1 million, and that was basically what we raised our money on. And then the next company, we went from like $1 to $15 million of ARR over like, I don't know, 2 or 2-ish years, 3 years. We're like, not a triple, triple, double, double even. And it was like the craziest ride. I was like thrilled going through it cuz I was like, we have product market fit. Like we can sell this thing. Like people want it, you know? So I think when you know the other side, it feels so good probably to be in this, in this seat now.

It feels, it feels really good. Um, I, I think I think that working at a hard tech company where we were not able to commercialize actually taught a couple features about maybe how to posit building a company that is a hard technology company, but getting some sort of product market fit. I do think it is, it's still kind of incredible. And I think, by the way, this is largely a result of the category of generative AI actually being valuable almost on day one, right? The moment ChatGPT landed, It was like valuable to consumers instantly, right? This is very not true about autonomous vehicles. I'll give you like a little, uh, like example here. In 2015, TechCrunch ran a story that said, this is going to be the year of AVs. And in 2024, they wrote a story that said, is this the year of AVs? Right? I don't think by the, by the end of 2022, which is a month after ChatGPT launched, probably TechCrunch wrote, this is the year of generative AI and it will be for the next 10 years. Yeah. Very different.

Yeah.

And, and Waymo was, it was founded in 2009, right?

It was founded in 2009.

Yeah.

The DARPA challenge was like 2003.

Yeah.

So it's like, it's 21 years of technology.

Yeah.

Yeah.

It is fun to see what's happening with you at the bullseye of this incredible time. Like we have this wave of AI that's like nothing we've ever seen. Then we have this use case, which is what Codium does, which is superpowering developers with code, which is all around structured text, which is perfect for AI. And then you have you and this team and this product that are like in the eye of this incredible hurricane. It's like what we, it's like what we dream about. Don't you think?

It is. You can't, you can't ask for much more as an investor. Um, I mean, I'm very optimistic that AI will, you know, kind of reach every area of our lives and every industry. But when you thought— think about where are the most sort of hair-on-fire problems and the sort of best use cases? Like coding is just one of maybe the top 3 use cases after like copywriting and then maybe like some sorts of customer support. Coding is like in the top 3 in terms of it being immediately useful now.

Can I revisit the conversation of like knowing how bad bad can be, what great looks like feels so much better. When you were at Scale, you were like what, employee?

I was around 6 or 7. I was an intern when we were 3, but then full-time. Yeah. 6 or 7.

And Scale's now whatever, like one of the hot AI, whatever it is.

Over 1,000 people.

Yeah. It's like a huge company.

Yeah.

Yeah.

Um, did you, did you have any, was it your first job?

It was, it was my first full-time job.

So did you have any idea? I mean, at the time it wasn't what it is now. Like, did you know how good it could be or what it, Was then?

Yeah, it was. It's so funny. So when I was in college, um, I mean, I was a math and CS major, so I didn't even know what venture capital was.

Yeah.

Um, nor did I know what a— like, what a SaaS benchmark was, or triple, triple, double, or any of that. Um, you know, but what I, what I did know is, you know, I knew that machine learning was something that I was really excited about. So even back then, in the sort of like computer vision days, it was like, I grew up doing math, competitive math, and then being in college, like, where is the coolest application of math? To me, that was even, you know, the ML classes at the time at MIT. And so I knew that's, that's like an area I wanted to work on. And then I knew Alex Wang was a special person. And obviously, I had not met, you know, the thousands of founders that I've met now. But I just got very lucky in that one of the very first founders that I got to know was Alex. And so ultimately, that's kind of what drove me to join. But I did not know when Alex made these insane revenue targets and just like, you know, a lot of the stuff that he did was just so out of the norm. I did not know that. Yeah. That's all, that's all that I knew. Sounds familiar. Yeah.

Yeah.

Yeah. Yeah. It turns out that, uh, yeah, maybe the, for the average company it would be very different, but for like the very top tier companies, I think there's, there's a bit of a pattern there. Yeah.

Yeah. How did you guys meet at MIT?

Yeah, we met at MIT.

Were you in the, like doing math team stuff too?

So I actually, I actually knew about Leigh Marie even before I went to MIT. This goes back to a thing of like a lot of us did math competitions. Yeah. In the past. Um, I think you made RSI. I did at the time.

Yeah.

Which is like this famous like research program, uh, that MIT Harvard runs for juniors.

What do you mean made it?

Uh, like they only select what, 50 students?

They do. Yeah. So they take like 50 high school students, um, their junior year and they bring them to MIT over the summer and we do research. So I did math research there.

Yeah.

I didn't get into that. She did.

Probably helps that I was from Alabama.

They try to do some amount of, you know, she's definitely trying to downplay this. I mean, she was the valedictorian of Phillips Exeter, right? So, I mean, she's, she's a big deal.

Yeah.

I had, I had known about her even before I went to MIT actually. Um, but I did a lot of the same math competitions, uh, probably like maybe some folks were in the same circle. In fact, even Alex Wang. So a lot of people actually that work in Codium, and we can get to this at some point, previously did math competitions as well. Like I knew them since high school, even since middle school. And our first employee, I went to high school with him. He was my roommate at MIT. He went to the physics camp with Alex Wang. And I think they may have even been roommates in physics camp. So like the connections to just the same set of people, it's a very small crowd of people actually.

That's crazy.

Yeah.

Do you guys, do you feel like when you, like now it's our business, when you see good people, you kind of like mentally slot them somewhere. and be like, hey, whatever it is, in some capacity, I wanna work with this person. Like now we kind of have to do that. But did you find yourself doing that then? Or did you feel like the bar at MIT was just so high that you're like, yeah, these are all like exceptional people?

Yeah, I think, you know, I think if I was to say the biggest privilege that I have is honestly just being born in Silicon Valley, right? Like a lot of the people that currently work at Codium, interestingly, are the original groups of people were all people that I just competed against in middle school and high school. And I would never have thought that they would be relevant in my life at this point, but they are just so incredibly talented. Like, this place is, is, uh, it's very cutthroat. It's extremely competitive, very cutthroat. And anyone that makes it out of here, uh, is, is probably gonna succeed in some capacity, you know. And, uh, and even at MIT, a lot of the people there were from the Bay Area or they did math competitions and I don't think we really thought about it in terms of, hey, I'm going to like recruit these people to come work for me. At that time, I don't even think I had deep aspirations to start a company at MIT. But no, it's honestly crazy how many of these people are still relevant in our lives. Like I talked to Douglas, my co-founder, all the time and I'm just like, is like, why is this happening? Why are the same people still relevant in our lives? Like, it's just like I'm going back 15 years. And these people are still like, just, I'm just looking at them right across the desk from me.

Yeah.

Yeah.

Yeah. I guess like you're probably not thinking when you're in the middle of a math competition, like, wow, we're going to build a huge tech company together. Right.

Dude, I didn't even know that tech companies were like a thing. I kind of knew Google was like an important company when I was in high school. But beyond that, and I kind of learned about MIT in high school, but I was pretty narrowly focused on School slash doing these competitions. I didn't really think about anything else.

Can I, um, can I play something forward to you and get maybe both of your reactions to this? So, um, Codium today is, well, there's stats everywhere on how efficient developers are getting or more efficient or what code is writing for them. Anywhere from 20 to 40% is like the range. Fair? Is that about fair?

Yeah, I think that's fair.

Do you think it's gonna get better from here?

Yeah.

Okay.

I think, I think, I think the ambition for the company is for it to get potentially 95+% more efficient. Like developers are going to be able to build application software potentially 20 times faster than if they didn't use Codium.

So let's take the conservative estimate of that. Let's say developers can build code 5 times faster.

Yeah.

Which we're probably not even far from today. Like what does that mean? For the world?

Yeah, I think that, and this is, you know, we'll see if I'm right. I think that the number of developers is only going to go up. Maybe the way they write software is going to be at a higher level. Like they're not gonna be writing the same kinds of programming languages. They're gonna be writing almost something that looks closer to English that is a little more structured than that. But the way I'd like to look at it is right now Codium sells to the world's largest companies. And if I was to go to any global CIO, a chief information officer, and I ask them, hey, I'm going to give you back efficiency, what do you do? Probably 99% of them will say, I need to get— I'm going to go and invest more in R&D. Because the way these businesses look at it is, let's say that you currently make $100 in revenue, right? The business can put in $20 into R&D. And the thought process is that $20 in R&D now yields $20 in profit. So next year they're going to get $120. Now, if that $20 in R&D is going to give them $40 of profit, just because R&D is now more efficient, They should be investing actually a larger fraction of their total revenue into R&D because it's a higher return on investment, right? So I think the point is actually if the cost of technology goes down, I actually think we are going to get not just a linear increase in the amount of technology, we're going to get a super linear increase in the amount of technology. And I think the number of developers and the people that are going to be contributing technology only goes up. And this is a unique aspect of technology, right? Like the world is never going to be satisfied with the amount of technology. that is getting produced. It's very different than food, right? Tomorrow, if, if Jubin, you ask me, hey, like, Varun, I can make food 10 times cheaper, I'm not going to say, hey, I'm going to consume 10 times the amount of calories. I'm just going to explode at that point, right? So I think, I think technology is a very, very unique aspect where I think there's, there's no limit at this point.

And, and just to continue to pull that thread, if people consume 10 times more technology, then what?

I think abundance in the world. I think consumers are gonna be very happy. I think GDP is going to increase substantially faster than it does right now. Um, I think, I think society's gonna be better off for it.

Yeah. What do you think?

I, I, I agree with Varun. I mean, I, the, the way that I kind of like to think about it, um, maybe this is an oversimplification, but rather than, you know, oh, we're just faster, so therefore there'll just be fewer engineers. I, I just think that's false. I think, We're faster, like we'll want to do a lot more. Like I think of, I was a product manager for quite a while. I think of all the tasks in the backlog, you know, at Scale AI that I would have totally done if I had had more like bandwidth with my engineering team. And theoretically, if they're, you know, twice as fast, even more as fast, I can get way more of those tasks done. And ultimately, I think software is just going to like be much, so much better across the board. There's so many problems with software today. It's not like perfect. Personalized. There's just a lot of low-hanging fruit across tons of different industries. I mean, we can go on and on about all the crazy large companies that still have, you know, pretty, pretty broken software experiences. So I'm just very optimistic on like us building more tech, writing more software, and just generally things, things being high quality, higher quality.

Yeah. But if you're like, let's just say like, like my, I don't know, like my, my mom. Okay. And you're like going to the grocery store. Why do you care if Safeway can produce 10 times more applications? Or if you're going to the gym, like, why do you care that Equinox now has developers that are supercharged in a way that they weren't before? Like, why will that actually make a difference to your life?

Yeah, I think the way software plays itself in a company like Safeway is maybe multifold. So let's take the first way. The first way is maybe they can more efficiently show people the goods that they want to buy. So shoppers have a better experience inside the store. That doesn't feel like incredibly sort of innovative, but maybe that's like step 1. Step 2, and this is, this is me projecting a little bit more forward. You know, when you look at the different kinds of software out there, you know, the biggest, the biggest kind of software that we have yet to crack that I think is going to be transformational in the world is, is physical manipulation, right? Like think about robotics. At some point, Safeway will probably have some bots in the store that are programmed by software that, that can automatically sort of put goods or food inside your cart and you just kind of leave Safeway and a robot just gives you the goods directly from the store potentially. Right. And I think, I think imagine a world in which, and this is kind of crazy, you like, you know, let's say, let's say you need to paint a neighborhood. You just write a piece of code that says for house in neighborhood, paint house. And I'm simplifying a ton, but think about how crazy the world could be. If you could actually almost do these kinds of operations. And this is what I'm trying to say, right? Like things that you basically can't even imagine today are potentially going to be possible, right? It's not going to fit the mold of what we saw in the past. It's kind of like how when the internet came out, people were probably like, how is this going to affect banking? Right. And the reality was most people were writing checks, going in person, and now they do everything online or they now do it on their phone. The modality has completely changed.

Yeah. That makes total sense. Um, can I ask you, like, uh, going back to this idea of being in the eye of the hurricane, um, the biggest team you've— like, what was the biggest team you'd like ever managed?

At Neuro, the biggest team was probably around 15 to 20 people.

Okay. And so now the company's what?

It's over 100 people. So probably close to 110.

And it's— we're gonna what, double it next year?

Probably more than that.

Probably more than that.

Yeah.

So like, uh, like, uh, how does that work? Like, how do you like talk about the eye of the hurricane?

Yeah.

Like you're so far, you've, this tech has never been built before. This startup has never been seen in its current form before. Uh, this growth has never been seen before. You've never seen this team size before. Like, uh, how does that feel?

It feels scary, but also at the same time, the way we're growing the company it feels, it feels not so scary. And I'll maybe give you an example. I think like we've, we've run the company in kind of an interesting way. I don't know if we're gonna talk about this, but we were, we actually pivoted the company at some point to even get to Codium. And we pivoted the company after making a couple million dollars in ARR. We were, we were at a reasonable amount of ARR, not something that's like a venture-style outcome.

Right.

And, and at that point we never hired a sales rep. And the reason why we never hired a sales rep for our previous product was because I didn't know how to increase the amount of sales. So I think this leads to a little bit of a mentality of the company, which is that we don't go hire positions at the company unless we're almost dehydrated. I think of, of the organization as if it should be this dehydrated kind of organization that is looking for water, and every employee is, is some, is like water to the company, right? And we don't hire someone until we're incredibly dehydrated. And then the moment we hire that person, we're dehydrated again, right? So it's, it's almost that every person we're adding, they get saturated the moment they join the company. And as long as we run the company in such a way where our ambition is large, naturally speaking, if you have a large ambition, you need more people, right? I don't think you can, you can say, hey, we're going to do something transformational for the world and be a 10-person company. That's, that seems a little bit faulty, right? But as long as you run the company in such a way where, as long as every person that joins the company is going to be underwater the day they join, it's probably going to be fine. Like, I think the biggest problem with larger companies is that People join the company for poor reasons. And I think most people are good. They're not, you know, kind of like evil people, but in an environment where they don't have much to do, they're going to optimize for things that don't really matter. Right. And then they start playing politics. They start having bureaucracy. They start hiring more people unnecessarily for their team to convince other people that their teams are important. And as long as we keep like the main thing, the main thing, I think we're gonna be fine.

When you— are you over this idea of like, I'm in venture, I'm no longer an operator? Meaning like, when you see this company growing as fast as it is, are you like, hey, that like, we got to let this hand play out? Like, there's nothing I can— like, I can't do anything. Like, this is— this thing's going. I don't know, like, do you stay up at night? Thinking like, holy shit, like, are we going too fast? Or have you found solace in the idea that you are riding shotgun at best? Maybe you're in the backseat of the car, you know, like, you're not— your hand's not on the steering wheel right now. Have you found solace in that?

Yeah, it's a lot of interesting things to say about that question. It's interesting. I mean, there's many different schools of thought in venture when it comes to, like, how impactful is an investor, how impactful should an investor be. And it's interesting, there are some firms that basically say, you give the founder the money, and you hopefully pick the very, like, the best people that are going to win, and you do no harm. And that is the job of the investor. And then there are some firms that, you know, literally employ hundreds of people, if not more, to come in and then help the company after investing. I'd say I align probably with the Kleiner school of thought, which is good, which is, There are things that we can be quite helpful with, but ultimately you are running this company, and that's why we invested in you as a founder. And so certainly, like, it can be a little nerve-wracking given how fast things are moving, like how dynamic the whole space is, all the competition. But from day one, I've always said this, I go to sleep at night peacefully because I know Varun is in charge of the company, and I trust what the decisions that he's making. And so though we'll, you know, we're constantly chatting on the phone, you know, you and others at KP have been super impactful for Codium when it comes to like the customer intros, the like market work that we've done, like building the sales team to the degree that it's gotten built in a year. But, you know, I like that. I think that we've been helpful and positive, but at the same time, ultimately, like the person and the team that is responsible is Varun and Codium.

Yeah.

Um, yeah, I think it's a careful balance and, you know, I think that I, I'm, I'm glad that Verena and I have the relationship that we do where we can probably be pretty honest with each other. Like what's, what's good and what's bad.

And yeah.

Yeah. How to, how to—

Do you agree with that take?

I, I agree with it completely. I, I think, I think maybe one thing about Leigh Marie in particular is like, I respect her opinion a lot. Uh, like I think a lot of people in venture capital, likely speaking, don't really understand the technology, right? They don't see where the puck is going. And they're trying to play maybe a pattern recognition game. And I think she's able to see beyond that. And part of it is also, you know, not only is she very smart, I think also being at scale, scale is like an anti-pattern to basically every company.

Hmm.

I think this company is constantly reinventing itself. I think we have had to reinvent ourselves once. I expect ourselves to have to reinvent ourselves again, right? I tell everyone at the company, expect something, shit to hit the fan at some point.

Hmm.

And that'll happen over and over and over again. And I think the fact that I know that she's seen that she's like very intelligent means that I can have like very honest conversations about, hey, maybe we need to change the direction of X completely. Right. And, and that's like very comforting to me.

Yeah.

Yeah.

Do you have a school of thought on this, this tension of like venture firms should do nothing, which is actually like where Leigh Marie came from, was that that was the posture of the firm. to, hey, venture firms can actually help. Like, do you have a— do you fall anywhere on that linear camp?

Yeah, I actually think so. I think maybe let me put it this way. I think venture firms can be incredibly helpful. I think they can be also very harmful. And luckily speaking, all the investors we've had have been like very, very helpful. I will say like the biggest thing is when we actually pivoted the company, we pivoted after raising a Series A. So, we had already raised $28 million in revenue, $28 million of capital, right? And we're at a couple million in revenue and we're telling our Series A investor, hey, we're just going to torpedo the entire thing and we're starting from square one again, right? And by the way, at that point, the fact that they were just like Green Oaks in particular, they were just willing to just say, hey guys, just do what you think makes the most sense, was very freeing, right? Because for us, it would've been terrible, honestly. And by the way, I hit myself over the head over this. Should we have pivoted 3 months earlier? I always ask Douglas this, my co-founder, should we have done this 3 months earlier? Because every month matters. Every month is very, very important. The fact that I didn't need to worry about that and we could just go with our instinct was huge, right?

Mm-hmm.

But then also working with you all at Kleiner, I've also gotten to see, hey, like maybe actually there's a way to accelerate the business substantially faster, right? Maybe there is a way to do all these things and you guys have been incredibly helpful to us And, and yeah, I think you have been incredibly additive to actually our experience. This has been more than just, hey, letting us go our own way, right? Being a thought partner, being actually able to introduce us to customers and partner with customers early on in the lifecycle of the company, I think has been incredibly helpful.

When, when you were pivoting, when you were pivoting the company, at any point did it cross your mind that this was a bad idea?

You know, I think the answer is no. And the reason is because we had one path, which is certain death, but slow. And the other path, which was high probability death, but fast.

But was it actually certain death at that point?

You had a couple more months. Yeah, I do want to emphasize one thing, which cracks me up. You know, the way that Varun talks about it, you would think, oh my God, this company's been around for like 7 years and they've made like a few million in revenue. No, it was like, If you did the— if you just did the naive, like, graph of the revenue, they were, like, best in class.

They're, like, on the path to a triple, triple, double, double.

But, but I mean, and this is— and seeing that, like, if you were just kind of playing the founder game for, you know, the vibes, or just, like, you know, just listening to what the average person would tell you about that, you'd be like, oh, I'm doing great. I should, like, pat myself on the back, and we should keep going, and we should take the $28 million and just, like, keep doing the same thing. And that's just not who Varun is or who the team is. And I mean, that, that's why I'm so confident that like, if something does come up, if there is a new tech development, a new competitor, whatever, I'm very confident they will, they will do it again. Yeah. If they need to do it. Yeah.

Like it wasn't like you can't— maybe you're looking backwards through rose-colored lenses, but you're telling me at that moment in time, after raising a Series A from a legitimate venture firm, with a pretty big Series A, mind you, doing something that was actually solving customers' problems, that it was an easy decision?

I guess the big issue was maybe this is where I try to be in like sort of a student of the game. So one of the people, our early sort of angels, was this guy named Carlos de la Torre, right? And he's the CRO of MongoDB. He's kind of telling me the story of how enterprise sales kind of grows. And at this point, you know, there's this famous book by John McMahon called, you know, The Qualified Sales Leader. And I'm reading this book, right? And we never hired a sales rep. So I'm the seller at this point at the company. I'm just like, there's no way this scales, right? We're selling to a handful of customers. I don't know how to even 3x this or 5x this in the next year, right? We can't even hire a seller because by the way, this is the other thing. I don't, you know, the company's DNA is not to go hire someone if we don't know how to do it. We hire like good people. If those good people can't get us 1% of the way there, we don't just hire someone to get us 100% of the way there.

Mm-hmm.

Right? So, we can't hire anyone. We're kind of stuck. We don't have a lot of great signal on what to work on next. Maybe we can eke out like, you know, $1 million here, $1 million there. But how is this thing going to scale to a business that does $50, $100, $200, $1 billion in revenue? And there were a couple of other things that also kind of bothered me a ton, which was actually We pivoted in the middle of 2022, which was when sort of the GPT-3.5 model started to become very, very popular, right? And that model actually gave us a lot of confidence that our business made very little sense. And, and the reason is because we saw that the transformer was getting very popular. And if one model architecture, which underlies the GPT models, gets very popular, what are we doing by providing infrastructure for GPU systems, right? We were basically going to provide a commodity. It's like asking, hey, Jubin, how cheaply can you run 1+ 1? You'd say some number, I'd say some number, and it would be a race to the bottom. And both of those factors in conjunction were enough for me to say, hey, I don't actually know how we can do something differentially better for customers. And I guess, you know, once you convince yourself that's true, there's no way to unconvince yourself, right? Like, we have smart people at the company. I guess that's another thing. We have people that I think are much smarter than I am, and it's hard for me to lie to them. It's hard for me to tell them, hey, like, you know, this business is awesome, keep working on this problem. Because we work very hard. We're an in-person company. People are at the company like late at night. I can't continue to motivate people. I think I'm like a pretty, pretty sort of convincing person, but if I don't believe it, I can't motivate someone.

How big was the company when you pivoted?

So we were actually 8 people. So we were free cash flow positive at that point. So we were making more money than we were, We were burning.

And did all 8 stay through the pivot?

Yes, all of them.

And how long before, from the moment of pivot to when V1 or alpha, whatever, of Codium was released?

I think it was probably 2 months.

2 months?

2 months.

Which feels very fast to me.

Yeah, I mean, we needed to, otherwise we'd lose morale at the company.

And, and then how long after those 2 months did you you close, like, did you close your first deal or did you start to get a sense that maybe we have something?

No. In fact, it was actually, it was actually really sad. It was really sad. So, so maybe at the time it didn't feel really sad, but you know, we were getting users. We never got users very quickly, right?

We were giving away the product.

We were giving away the product for free. So this was an interesting thing that we did, which was that, hey, instead of being an obscure company, what if we became a company that more people knew about? Let's not focus on value capture. Let's focus on value creation, right? So we, we created this individual product, Codium. We gave it out for free. The reason why we were able to give it out for free is we were originally an infrastructure company. So we were able to run really, really efficient infrastructure to go out and give the product away for free without burning a, like a lot of cash, right? So we never got users that quickly. We did have enterprises sort of start to reach out to us in January or February, right? So the launch is in November. We have enterprises reaching out to us in January or February. And then one of our big competitors at the time is GitHub, right? They have this product out called GitHub Copilot. And in March, they release this thing called Copilot X. Okay? And this is like a vision statement from Microsoft. You wouldn't have expected Microsoft to come out with a vision statement, but it's a bunch of features of what they're gonna do. They have yet to build 95% of what's in Copilot X, and it's been over a year and a half. But at that time, all these enterprises get spooked. They're like, hey, why would we go work with Codium? if Microsoft is saying they can do Copilot X? The reality is they could never do that. And the morale hit internally at the company was so much more, it was so much higher because we could not believe that Microsoft would be making a vision statement. We thought any, everything that they said they would do were gonna be things that they actually did. And, uh, yeah, it took us many months after that before we actually closed our first enterprise deal. But the morale hit of the company is all like, was very up and down. It was like, hey, we got some users. Hey, an enterprise reached out. Hey, all the enterprises are pulling out. Hey, we're getting more users. An enterprise came in. An enterprise actually closed. A good enterprise that we thought was going to close didn't close. And then it's always been this kind of roller coaster ride, which has actually been very good for the company. Because I think people have— people work very hard, but they are not deterred by bad news, right? When bad news happens, the first thing all the employees at Codium want to do is act and do something to kind of do better. Right? Not like get very demoralized. So, I actually think the pivot and all of these things were the best things that could ever happen in the company. It's forever changed the culture of the company to be this like super truth-seeking company that is willing to move really, really quickly.

How many competitors do you think you have today? When I say competitors, I mean people that are attempting to do this, have raised a ton of money.

Yeah.

How many do you think there are?

You know, the number of people that are direct competitors, maybe I can count on one hand. Um, at this point, but there's like a lot of these companies that kind of have raised a lot of cash without a product.

Hundreds of millions that have never released anything.

Never released a product. This isn't uncommon though. This is not very uncommon. I sort of worked in autonomous vehicles where all the companies were raising billions of dollars and none of them have yet to ship a product. And in fact, it's very, it's very like satisfying for me to see that because for me, what I'm thinking is, what are their employees thinking? How are their employees thinking they're winning? Right? They're seeing other companies that are, that are satisfying customers, that are scaling the number of users. Their employees must be so demoralized. And that gives me a lot of solace every day.

And Leigh Marie, for you, as you like, we looked at all the companies. The challenge with making a bet in the space is that's our bet in the space.

Correct.

And going back to my earlier point of like, this is the mother of all markets in an incredible space that's perfectly tuned for AI. Like, how did you make this one?

It's, it's, it's a great question. And I think, you know, one of the challenges of being a true multi-stage generalist investor is that you can theoretically invest in any private company at any time. And so you can pick a seed early on, you can say, hey, this is our bet in the space, we really like this founder in the space, let's just, let's do it at the seed. Or we can wait and we can get sort of more Traction, more data.

Yep.

You know, there's no guarantee that you aren't gonna have to pay some sort of crazy high price, or there's no guarantee that you'll even be able to get in the company after it becomes consensus enough. So it's constantly something that, you know, I'm thinking about, especially when there's a great space like this. And so when I started actually at KP, this was in mid-2023. So this was, you know, months after they pivoted, they had started to get some customers, especially towards end of 2023, and a lot of momentum. And so basically when I started at KP, the thing that I was working on the most was cogin, just mapping out the cogin market and figuring out like, yeah, who should we partner with?

Because you love the space.

I love the space. I mean, and it's, I love the space because I think there's the opportunity to build a giant business. Personally though, I also love the space given that it's AI and engineering, 2 things that, you know, are just kind of like very core to who I am. And so when I got to KP, I just started doing a lot of prepared mind work, making sure that I had met every company, knew everyone's strategy, knew everyone's sort of like founder quality potential. And then it just, it became very clear after I just like started this doc with all these different companies, like, no, the correct opportunity is just to get into Codium as fast as possible. And thankfully, I had a team that was very supportive of us doing that. So, I mean, I'm happy to chat through the thesis of the Codium or, yeah, whatever.

Well, I guess like at the time it was not that, like it was early. Yeah. It was so early. So there was still no data really. There was like, there was probably less revenue in this version of Codium than ExaFunction had done.

Yeah.

At the pivot.

There was.

Right.

Yeah.

So it was so early. So really at the time it was just Varun, the team, and then the market, right?

I agree with that somewhat. I do actually think though that we had some sort of like asymmetric information in that, I mean, and you know this very well, Jubin, but we're just very involved with CIOs here, right? We have multiple CIO networks that we're a part of. And I think the sort of value prop that Codium is going to be this more secure, more context-aware, more sort of like extensible, like will play nicely with any dev tool stack. That sort of value prop, maybe at first glance is like, oh, well, you know, no, probably everyone just like uses GitHub and Microsoft for their whole entire codebase. And, you know, why would they need any of that? And it's just so untrue. Like the average enterprise has a very fragmented set of dev tools that they use, have come up to problems in terms of the sort of contextual awareness of certain code assistants. And it just became clear talking to CIOs that these were real problems. And so having that sort of like data from, from the CIOs, plus like the usage of Codium in particular, we did get some usage numbers before we invested, was off the charts. Like, I think, you know, my partner, Moomin, compared it a bit to Figma, just in terms of how much are people using this product and loving it.

Yeah.

So we did have some amount of data. So like, we certainly— there was like a bet that we were making kind of on the momentum. I think it's kind of surpassed our wildest dreams in terms of what it did. But there is, there's some amount of data.

What's, what's really cool to see, and we're still talking about this year, right? Like we're all still, this is still like from the beginning of this year till now. I remember at the beginning of the year when we would talk, it was very much about like, hey, we think the early customers are gonna be security-conscious customers, people in the defense industry, JPMorgan Chase. They did end up becoming customers. That's interesting signal. But then like very quickly, not banks and defense departments were signing on to be customers of Codium. And then I think you like, it widens the aperture of what that opportunity could really be, which just made us more excited. It's really crazy. It's amazing that this has only been this year. Varun, can I ask you, like you had mentioned, like, hey, we're gonna go build a, like we pivoted because it's not gonna get us to a $500 million or $500 million revenue, billion revenue company, the previous version. Why do you care? Like, why does that even matter to you building a company of that scale?

You know, maybe the reason is like a little bit of the motivation for doing the company in the first place. Like, we hire actually amazing people. I think these are people like over 35 to 40% of our engineering team is from MIT. But I think even That the people from MIT, I think, are like the top 3 to 5% of all of MIT that are coming to the company. And I think the reality is if we don't build something great, these people have no reason to stay at the company. And I want to work with great people. And if I can't work with great people, I'm not that interested in like a middling outcome for the company. I'd much rather, I guess, work at another sort of innovative company that I could build something crazy. I tell people this all the time, but if I was able to work at Starlink, you know, for at SpaceX, and I was like one of the first like 10 people there, and I was able to give millions of people internet, I would, I would do that. I wouldn't like start a company. Like, my motivation has always been, can I be a meaningful piece, a part of like a visionary company? And if we're gonna be like B2B SaaS XYZ software, I don't think I'm like that interested in that.

Does it get old? I asked you this question when we were going for a walk, because you were kind of like, how would I describe it? You were very keen on understanding how quickly are you growing relative to other people? How are competitors doing? You were very just concerned about everything that was going on. And I'm like, this is just how he is. Like, he's just always on it with every single detail all the time, like relentlessly on it. And I think I asked you like, Does this get exhausting? Like, does this feeling of being so almost neurotic about every detail in every part of the business in the market, does that like tire you out?

You know, I think actually the pivot and actually has made me more, more anxious because basically what this, what it has shown is with new information, if we move very quickly, we get to the right answer faster. So actually, you know, my, my takeaway of all of these things has not been, hey, maybe it's fine to sweat the small things and not know a better thing that we could be doing for a little longer. It's actually made me more, have more urgency. And I think the nature of the space, given how quickly everything is moving, like we just cannot expect to be the same product every year. Like every year we should almost be cannibalizing the previous year's product. We should almost make the previous year's product seem silly. And if we're not on a path to doing that, I think we're on a path to death. Like, maybe the right way to put it is, you know, I respect Microsoft a lot, right? They have like more resources than us, more capital than us, more distribution than us. And I think we can innovate substantially faster than them, right? I think a lot of the people that joined us would not want to go work at Microsoft. But I think if we take our foot off the gas and we are not innovating very quickly, they will very quickly commoditize our technology. It's not really like they might have people that are, that move less quickly than we do, but with enough time they will find a way to commoditize the technology. And I think I'm perennially worried about, are we moving at the speed we should be moving? And honestly, are we moving fast enough on product? Are we moving fast enough on distribution? Right. The way I like to think about it is product defines the ceiling of the company and go-to-market defines like the exponential curve you take to get to that ceiling. And it's constantly this game where we're like trying to increase the R-value, the exponential rate of that exponential curve on go-to-market, and then also changing the ceiling of the company. And if we're not doing changing the ceiling of the company fast enough, the ceiling is also coming down because the technology is going to get commoditized quickly. I think there is a good deal of paranoia for everyone at the company, right? A healthy amount of paranoia for everyone at the company. And yeah, like we just signed up for it working at a startup. We're not at Google. So it is more painful to work at a startup than it is to— our company than it is at Google. And we've kind of just accepted that. And I guess I've accepted that too.

Can you take us into— you're releasing a new product in the next couple of weeks. By the time this airs, it'll be out. Can you take us into the walls of Codium on Castro Street? Is it on Castro?

It is on Castro Street in Mountain View. In Mountain View.

Yeah, yeah, yeah.

Not SF Castro Street.

Yeah, yeah, in Mountain View.

Uh, I'm a big South Bay guy.

You're a big South Bay guy.

Yeah.

Can you take us into the walls of Codium? Like last, last 2 weeks sprinting on this product, what's it feel like inside the company?

I mean, it feels awesome. Like one of the things that I was really worried about is as the company got bigger, would people not feel as much passion to go and work very hard, right? Because our company at this point since the beginning of the year has over 5x'd, probably close to 6 or 7x'd in size. Right. And engineers are at the office until 3 or 4 AM to make sure we're able to hit this deadline. Right. It's self-imposed.

Right.

I'm not out here asking everyone like, you have to be here at this time.

But are you at the office till 3 or 4 AM?

I'm at the office pretty late too. Yeah. Yeah. I mean, it would be weird if I wasn't and everyone else was working very hard at the same time. But no, it's, it's people. And this actually comes down to what Leigh Marie said. We picked a problem and this is the important thing. The problem that you decide to work on actually is very important. we picked a problem that engineers are very passionate about. Engineers are passionate about how software could be revolutionized, right? The way to create software could be revolutionized. So everyone at the company is a missionary, right? They're not only brilliant, they're also missionaries for the product they're building. And I think the last couple of weeks we've decided to go ahead and build this IDE called Windsurf. And I think it's going to be really awesome in that it's going to let the creativity of the engineering team to have no ceiling, right? So far we've been playing in a couple of these walled gardens, a couple of these other IDEs that— and IDEs are the platform that developers write software. And I think we'll, like, the ceiling of creativity for the team is going to be uncapped, right? And it's exciting to see the team working this hard. And if you think about it, the original sort of group of people at the company, they kind of built the product without us even knowing how we'd make a dollar of revenue, right? So the motivation has always been we want to find a way to completely transform the way in which software gets built. And along the way, we realized we could build an enterprise business that could be like one of the world's largest enterprise businesses, hopefully. But it'll be great to sort of go back to kind of basics on this.

Can you tell the story of how the Dell partnership came together?

Yeah. So I guess the crazy story. And so we got an inbound message from Dell middle of last year, might've been actually it's early last year, early last year, 2023. And we're working with the team. We're trying to see how can we get Codium to run internally at Dell, right? For their developers. And we're doing this, but they're unable to get a GPU machine because our product actually runs on GPUs, right? We need to use a lot of deep learning to actually make the product run. And we're unable to acquire a GPU. And in the meantime, for some reason or another, I'm talking to Bipul, who's the CEO of Rubrik. I'm just, I'm just messaging him. I cold messaged him on LinkedIn. I'd never met him in person. And largely speaking, the only reason I asked him was Codium can run on-prem, can run on-prem or in a self-hosted way. And I just wanted to understand, he's, he's built one of the largest, I guess, on-prem businesses that are recently created, right? Rubrik has recently gone public. I just wanted to get a couple pieces of advice from him. And I talked to him on the phone and he gives me a lot of useful advice. Some of that advice I still tell the team today, which is very generous of him. I don't know why he even bothered. He has no ownership stake in the company. He didn't benefit at all, were meaningless to him in, you know, in the, in like the rational sense. And after that, he was like, oh, it seems like you're having some problems with Dell. Let me go email, email a couple of folks. I'm like, okay. I mean, this, this, you know, that's, that's great that he said that. A lot of people say they're going to email a couple of folks. The next day arrives, it's Saturday. And Saturday morning, I'm on an email thread with Michael Dell himself. Right. He's just like, oh, we'll get you settled out, Varun. Like, hey, know, here's a box, like here's a team, here's our president of sales. And just immediately the tone of this conversation has completely changed, right? Like Dell is now just able to get up and running almost instantly. And I think the cool part about the story is not only that Dell is like one of our biggest partners and customers right now, I think the coolest part of the story is how these moments of kindness, right? There's a lot of things that were random in the Valley of how people have helped us where there was no reason for them to, or there was nothing we could do to return the favor, right? Because our company made very little revenue or we had very little traction. And just very, very grateful that we live in almost Silicon Valley where everyone wants to pass it forward, right?

Yeah, I totally agree. Bipul was here, by the way, 2 weeks ago, speaking to a group of CIOs that we bring together a couple times a year. And I asked him, Why'd you do it? And, um, he said, because I felt like I, I owed it back to the system that kind of like raised him. And that was just a really cool answer. Like, that was just a really cool answer. And he knew he didn't need to do it. And he knew he did it because he probably saw himself in you and saw you as kind of the next up-and-comer that he felt like he owed. And going back to your point of like how special this place is, like that doesn't just happen. It's amazing, isn't it?

It's actually miraculous. And by the way, like I also take, you know, we're not— I genuinely try to, you know, tell people we're not like a meaningful company yet, right? We need to prove that we are a company to reckon with. We're so early in our journey. So I don't think we should get very confident about where we are. Everything we have could disappear at any point. I tell everyone this at the company. But even whenever I can, if someone's like, hey, I want to go start a company, I will try to take a while, potentially hour, hour and a half to like talk to them about, you know, their idea, you know, what they could do well. And like ultimately there is a little bit of, and this is where I try not to be too prescriptive, which is that every company's different. I'm not going to understand their business better than them. But actually taking the time to understand how I can be the most helpful because this is like a hard journey and And, uh, and it's nice to feel like you have a helping hand along the way.

Leigh Marie, do you ever— like, with Varun, like, we, we both know him pretty well now. Like, there's no— I don't think it's possible to increase or push his ambition.

Do you?

Like, like, do you— like, how do you think about your relationship to his ambition? Because In a lot of cases, folks will say, hey, where venture capitalists can be particularly effective is raising the bar on a person and a company's ambition that they can't even see. But with Faroun, in my mind, I've always felt like, I don't know, like his ambition is limitless. I don't know. Do you have a reaction to that?

I mean, the ambition is totally there. I don't think I can raise the bar on Faroun's ambition. I do think, though, that, you know, I can offer a different vantage point than Varun has. I mean, basically, since I've operated for, you know, years at scale, I think VCs have a very unique vantage point into the whole, like, Silicon Valley ecosystem. And, you know, many people in Silicon Valley, and this is one of the reasons why I love it, many of us are extremely ambitious. Like, we say all the right things when it comes to, you know, building big businesses or things like that. But then, you know, you have to live it out, and you really have to like come up with not just the right sort of work ethic and team and attitude, but also the right strategy. And so I think that's, that's really kind of the only way that maybe I can even try to, try to be helpful is in some of the strategy and in some of, I guess, KP's just unique advantages when it comes to relationships.

Yeah.

But I just want to add too that both all the answers to these questions, they kind of sound like an ad for Silicon Valley and working in person because, yeah, I think if you're a founder and you're thinking about like, where do I have the best chances? The ecosystem here, how much pay it forward there is, like how much you're just, anywhere you go, whether it's in SF or in Mountain View, you might run into somebody that becomes a really important part of your company's story. I just don't think you get that anywhere else. And I also think that to get the full benefits of that, working in person, like Varun told the story about him staying up at 3 and 4 AM with a company. I have a very similar story. Alex Wang would stay in the office all night and label tasks. If we did not hit our SLA, for customers in the early days of Scale, Alex would be labeling. And so I was labeling. So I would code during the day, and then I would label during the night, all of these tasks. And there's just no way if we were remote, and I was not next to Alex doing that, that I would have done that. And, you know, was that essential to Scale's success? I think so. Like, I think that if we had lost some of those customers in the early days, because we didn't have like all the taskers that we needed, it wouldn't have— it would have not made it. So. This is my ad to build your company. I'm very biased in Silicon Valley in person.

I, uh, is it unfair of me when people want to build technology companies, whether as a founder or an operator, but they refuse to move to San Francisco or be in the Bay Area? And I'm kind of like, I question their ambition. Is that unfair?

I think it's, I think it's, you know what? I think if you're building a real technology company, I think it's hard. You are making, you're stacking the odds against yourself. Right. Because just think about what, what the Bay Area really is. It's a bunch of people from a bunch of places that are selecting for, they want to work on technology, right? Like people that really want to work in technology and, and not trying to like diminish Ohio, they don't decide to go like, you know, settle completely in Ohio, right? That's not gonna be where all the technology talent is going to be. Yes, you need to pay a premium for people if they are in Silicon Valley, but hey, if you're starting a company, your odds of success are close to 0.1% when you start anyways. So why are you stacking the odds against yourself? Right? I try to treat it as, hey, like, look, I'm trying to basically run through fire right now. Why do I want to give up a leg right now? Right? And, and I'd rather have all my legs. And then, and if having, if having these super expensive legs and I can't scale at some point is a problem, then we'll deal with it then. Right? But, but it's, you know, it's a, it's a little bit of a premature optimization to be like, I can hire way more people because I just, I just think You know, a lot of people try to view building software and building technology like it's building a factory. It's kind of not. When you hire great people, it's like, you know, let— there's this classic meme of like having a 10x engineer. It's kind of real. Yeah. But the thing that the meme doesn't actually capture is when you hire someone that's 10x better, you can't just replace them with 10 1x people, right? Because the 10x person is gonna be thinking of ideas that the— that none of these 1x people talking to each other are ever gonna think of. They're never gonna come up with it. So optionality in the company disappears. And if you're running a company where the entire company can be run as a, as solving tasks on a Jira task board, then fine. Like go and build your remote company that sells B2B, like telephone text messaging software. That's fine. But like, if you wanna build a company where like you don't know the answers to questions and you wanna hire people that are very ambitious, motivated to solve problems, like you gotta do it here.

How quickly, maybe I'll pose this question to both of you. How long do you think it would take for you to figure out if someone is a 10x engineer?

I think it's pretty fast. It's very fast. Like, I think, you know, one of the things that, you know, Leigh Marie and I probably feel is like, we know what the difference between good and not good is. I think a lot of people are really bad at knowing what good is, right? I've just seen monsters. I've seen people that can process information many times faster than I can, uh, at MIT. And I can just immediately tell if I talk to someone, this person has something very, very special. And, you know, when I look at a lot of startups and I'll be very honest in Silicon Valley, I just don't think that they're staffed that well. Like, I think most big companies have more talent density than most startups in Silicon Valley. That's just because, you know, most startups are chasing people that were from X company or from Y company. Most people that leave from X or Y company are the people that are the worst from that company.

Right.

So, it's actually, you know, I think Spotting talent is a very, very hard problem, but like you have to have seen talent before to actually even spot the talent. And I will say also Silicon Valley has the best talent too, or we do have great talent. And I'm not saying, by the way, that the rest of the country doesn't have great talent, but if you want to build an in-person company from day one, it's going to be hard. Like you're just, you're just setting yourself up for many, many disadvantages in the very beginning.

Do you disagree with this take? You shrugged your shoulders like maybe you disagreed with part of his take.

I, uh, let's see, what was the, what was the part that I, that I disagreed with? Oh, the shittiest people leave big companies. Yeah, I guess, I guess no.

My point is the people you are getting that are very quick to leave a company are usually speaking not the best people from the company.

Yeah.

So you can't just say I got a person with this logo, right? That's— you're conflating 2 pieces of information, right? I talked to people that are from the best companies and I'm just like, are you worried about these other companies taking your talent? They're like, no, the people that leave are usually the worst people that are usually leaving my company because those companies know how to retain their best people.

Yeah, I think I describe it kind of in terms of like, I think there's 2 things, vintages and then appropriateness of the logo. So I think depending on the year, let's take Stripe for example. Amazing logo, amazing eng culture, really great engineers. I have friends at Stripe that have convinced me that depending on what year they left Stripe, you can like make a judgment about like how good of an engineer they are. Because I think there are just certain time periods where You know, there's, and there's time periods for every hypergrowth company where kind of like the OGs who are usually quite good, or some of them are quite good, leave. Maybe because they want to start their own company or something like that. And there are time periods where, yeah, maybe you have people that are leaving for not as good of a reason. So there's that. And then 2, I think it's also tempting to just say, oh man, I got a person from Figma. Like, they must therefore be good. But if they are not sort of good in the way that your company needs them to be good, Especially like if it's not just for something like a certain type of engineering, then it could not actually be that helpful for your company.

Yeah. And on the, on the vintages point, are you saying like those that stayed at Stripe for either not long enough or way too long probably don't fit the bullseye of the vintage of someone that was able to rise through Stripe but also had the ambition to at some point go leave and do something else?

Right.

Right.

Yeah. Yeah.

I appreciate you both doing this. I really do. It's been like almost an hour and 10 minutes, if you believe it. Thank you. It's a thrill to just get to work with you. Like, I think I speak on, on all of our behalves. Like, it's just really fun. And I do think that there is something seductive about ambition. And when you see someone else that has huge ambition, it's— you feel it like it permeates. And it permeates to your team. It permeates through these walls at Kleiner Perkins. Like, we feel that ambition. And I've said this to you before, but John Doerr has a famous quote where when you see someone that convinces you that they have the ability to pull the future forward into their present, it's a very energizing feeling. And that's that feeling that I think you give us, is this feeling that you're quite convincingly going to pull the future forward. And We just— I just can't wait. I can't wait to see what you do. I really can't. It's amazing.

Well, really great to sort of be able to work with you guys. Honestly, it's been a roller coaster ride the last year. Um, and thanks for putting up with us.

Of course.

Uh, are you hiring?

We're hiring for every role. Yeah, we're hiring.

Every role.

Yeah, we're hiring for engineers. So if you're— if you want to work at a very high performance engineering company, come, come join Codium. We're hiring across the sales organization. We're hiring account executives, mid-market, commercial, enterprise, strategic account executives. We're hiring sales leaders. Yeah, like there are lots of companies that currently want to use Codium and we'd love to be great partners to them. So across the board on go-to-market, across the board on engineering, if you're excited about the opportunity here about rewriting software as we know it, come join Codium.

When you hear the word grit, what do you think of?

Being truth-seeking. When it's very, very hard to, when it's humiliating to, uh, actually changing course, when you know in your bones that it's gonna be a very, very tough decision and it's easier to not do it otherwise.

Faroun, Leigh Marie, thank you.

Thanks, Juvin. This was great.

Thank you.

Thanks a lot.