Jake Cooper. Railway.

Jake Cooper:

Most startup advice is overrated. I didn't really start this as a company. It was never like a fully intentional thing until it got to a point where like I had no job offers but, a few angel checks and I was like, all right, screw it. Like I can always go back. I work like sunup to sundown, Monday to Friday, and then I try and close my laptop until like Sunday afternoon. We say no to like 90% of stuff internally, right? We're just like, no, like as of right now we're not going to conferences or like as of right now we're not doing like this specific stuff because we say no, we say no, we say no. And then when it's inevitable, we said, Yes, like that's what the thing we're going to go in and do. Approach for talent is like, never be stingy with, talent, right? I'm evaluating comp like, you know, every half, it's probably moving every quarter. It like never even really crossed my mind to like have this in, person because remote allows us to pull the best people from literally every facet of the world, right? We don't use AI internally for any of the memos or things that we go in and put together, right? Nobody should have to read the, thing that, you have slopped out and just like pushed onto somebody else, right? Like, how do we go and, you know, get to like 80 people next year, right? Maybe $100 million, and then we're doing $500 million, maybe more in like revenue. You don't realize how fucking bad of a manager you are until like you get better at being a manager.

Nakul Mandan:

Jake Cooper told me something over dinner last night that defines how he's building Railway. He said if Stripe is doing something, if Cursor is doing something, he doesn't want to hear it. He solves the problem himself first. Only as a last resort, he checks what others have done. That one trait explains everything about Jake. It's why he's a solo founder, why the company is fully remote, why with only 40 people he's approaching $100 million in ARR, why there was no sales team until last year, and why a software company owns $90 million of its own hardware. Railway is one of the defining infrastructure companies of the AI era. But the more interesting story here is the operating system of the person building it and what happens to that operating system when the company goes from $100 million of ARR to $500 million and beyond. Knuckle up. Jake, welcome to the show.

Jake Cooper:

I'm excited to be here.

Nakul Mandan:

This is awesome, right? So most people would have found Railway in the last 18 months and would accordingly position the company in their minds as infra for Vibe coders. But the company has existed since 2021. What is the origin story that they might be missing if they just discovered Railway?

Jake Cooper:

Yeah, 20— 2020, actually. and so like the origin story is like we started out as a way to like provision databases right? I was like, hey, there's no good way to do like stateful application hosting or database hosting at that point in time. And so we started as a database and we've just like basically walked all the way up the stack, right? Like moving to like backends, right? And then like frontends and like, you know, AI-based applications now and everything else like that, right? And so I think the fact that like everybody's kind of discovering us is like, it's very interesting because we started like so deep and then we kind of walked our way back and then we've also gone back deep again with like a bunch of the data center stuff that we're building. So.

Nakul Mandan:

But in those years, like '21 to maybe early '25 or late '24, what were those years look— looking like in terms of revenue? It seemed like there was a lot of free tier layers, and at one point you were even losing $500K a month. Walk me through those year— those years.

Jake Cooper:

I mean, so like, you know, when you're starting a business and you're, an engineer who has never run a business before, right? Like, you're really focused on the product. And still to this day, I'm like core, focused on the product. I've just gotten better at like letting go of those sections and delegating them and then being like, okay, cool, right? but I wasn't in the early days, right? I was like, okay, cool, we gotta make a really, great product that people are gonna go and love and all this other stuff. And then so like, I think it was like kind of right around when we did our Series A or whatever, but we were burning like a ton of money or shortly thereafter or whatever. And yeah, like we, had a ton of like original developer love because people were just like, oh, this is awesome. It's like so trivially easy to just like spin up like database backend, like whatever at this point in time. And like, again, I was an engineer and like had no idea about like pricing and packaging. How are you going to go in and do this? You know, we, talked a little bit like about this last night at your house, right? Like, it's the highest leverage thing that you can ultimately, do, right? And so, you know, I looked at the balance sheet at one point, I was like, oh shit, like, this is not good, you know, right? and so we had to kind of like make some changes, right? We ultimately like removed our free tier, kind of like pulled things back until we were saying like, okay, listen, these are, these are the— like, we're servicing this many users and we can only service this many users like really, well, right? And like, I'm a big fan of like Jim Collins, The Hedgehog Principle, et cetera. Like, do only the things that you can be the best in the world at. So if you're do— if you're doing all this surface area, we had to like pull it back. And so just made the call that we're like, it was, it wasn't gonna work otherwise, right?

Nakul Mandan:

What was the change? You cut off free users completely, right? So now everybody is a paid user.

Jake Cooper:

Well, we, cut off free users originally, and then we spent like 2, 3 years building out our bare metal offering so that we could go back and offer that like free stuff for those users. And so it was like, you know, it was like free tier on, free tier off, right? And we waited like 2 or 3 years basically till we could guarantee that we were like, okay, we can service this, portion of users, right? With like this pricing and packaging, right? And then, you know, we have no plans to go in and remove that indefinitely because it, runs now on like a very, solid like foundational business model that is growing very, rapidly, right? Where we can go in and monetize, more and more and more as folks kind of like grow on the platform, you know?

Nakul Mandan:

So in those years, it seemed like not just pricing and packaging, but maybe product-wise, there was a lot of iteration. How close did it get? Like, were there moments where you were very close to just running out of money, or you were always managing that part well?

Jake Cooper:

I think we were like— I mean, it depends on like what you define as close, right? Like, you, know, like, I personally don't like when the runway is like below like 18 months or whatever, right? And so we're getting into like the single-digit months of stuff being like, hey, we gotta, we gotta fix this. you know, and somebody internally was pointing this out and they're like, you know, like, we're not going to tell you what you need to do about this, but like, you just won't have a business in like 6 or 7 months if you don't do something about this. And so it was like, yeah, sure. And we already knew it was on, the horizon. We were already going to go and do stuff about it. So it was just like, all right, great. That's what we got to go in and do.

Nakul Mandan:

Then what changed that made the chart go vertical? I guess part of the question is, was it, was it a product-related change? Was it a macro trend? Was it marketing and messaging? Was it pricing?

Jake Cooper:

Yeah, so I would, I would actually argue that like there's just been multiple kind of inflection points, right? And so is that there's that like initial like free tier, like, oh cool, this like growth is free and you're selling, you know, dollars for, you know, like what, 20 cents or something and be like, wow, this is awesome. What a great product. Dollars for 20 cents.

Nakul Mandan:

And that was the database product.

Jake Cooper:

yeah, this database in the backend, right? Like that, was it, right? and then so we were like, all right, cool. That does not— that math does not math. All right, so let's like make a, a business that actually goes and works, right? And so then users, you know, pull away, right? Sign up charge goes down. you know, you, You can literally see it on our chart where it goes like this and then it goes— for like an entire year of just like trough basically. Right. And then we round the bend basically, I don't know how to say, like late '22 or something like that. And then it's kind of like compounding, right? It's like doing the thing. the business is interesting in the sense that, well, it was until this year, kind of seasonal, right? Because it's like a lot of SMBs, a lot of hobbyists, et cetera. So, you know, summers kind of tend to lull. This is actually like one of the first years where we're just seeing like the signup chart continue to go through because people have figured out how to run agents in the backend, right? Like, and they're just like, they're firing off jobs from their phone, they're firing off jobs on like, you know, like background jobs in like Claude, or they're just like, you know, talking directly to their phone and stuff like that. So yeah, so we've built up like kind of a almost like compounding. And then I would say that like there's been a kind of inflection point as models have just gotten, better and better and better. And now you can just see like as the models, as the model releases get better, the chart gets more sheer and more sheer and more sheer. Right? So it's just been really interesting. And I think like we're going to see another one of these like almost like step function changes in, the chart going like more and more and more vertical as everybody just kind of gets acquainted with like cloud agents, right? And just being able to get everything off your laptop.

Nakul Mandan:

I'm curious though, was it that the segment of users that suddenly got empowered with Claude Code and such that suddenly made the chart go vertical, no seasonality and stuff? Like, was it Was it that you ended up with a segment that allowed you to explode, or did the product get better, or did it just compound to a point? And I see this with some of our— even other companies where some companies or products, they just need 2, 3 years of compounding, both product and messaging, and suddenly they're ready to explode. But they didn't change anything, it was just compounding.

Jake Cooper:

Yeah, I would say it's definitely more iterative than it is like some sort of like thing that I could give to people as like advice of like, oh, just do this and your chart will like go vertical, right? Like The just do this advice is like show up every week, tackle the biggest problem for, the company that you're getting after, and then figure out some way to, finish your week by moving that like brick forward and just do that for 52 weeks a year. And then just like you'll have moved 52 large problems forward and then you'll compound and compound and compound, whether it is your messaging or your positioning or like your product or like your sales, go-to-market operations, anything else like that. It's like, you know, we have a saying internally, you know, it's pulled from an old friend from Robinhood. it's brick by brick. How do you, how do you build, you know, an empire? So you build it brick by brick.

Nakul Mandan:

Is the product fundamentally different today than 2, 3 years ago? Did the product footprint evolve?

Jake Cooper:

Yeah, well, I think right now it's definitely evolving, right? Because we're seeing a lot of those like long-running asynchronous jobs that can fan off other jobs, right? And so we've got a few things in terms of like, okay, now we've got an agent in the dashboard and the CLI and the MCP, and you can, you can offload jobs to it. You can, you know, run ssh railway.new and get just like an immediate, like virtual machine that you can kind of write and run code on. You don't even need to be signed up, right? And so there's a a lot of like these kind of footprint things that are changing. And so we've kind of peeled apart the, a little bit of the product and then now we're kind of layering in a lot of those like almost agentic experiences inside of it.

Nakul Mandan:

Right. So one thing about you that I have noticed in our few conversations now is that you're almost allergic to knowing how other companies have solved the problem that you're facing. You, love solving your problem yourself. Walk me through that. Like what led to that philosophy overall?

Jake Cooper:

It's kind of, to, me, it's, it's almost like cheating on your homework when you go and look at like somebody else's solution. You like pollute your own, like, brain with whatever that is, right? And so I personally always try or like, you know, push people internally to be like, listen, if somebody else has done this, like, you know, if, this is how Stripe does this or this is how Cursor does this or this is how anybody else does this, like, I don't want to hear about it. I just don't want to go in and hear about it until we have a good reason to say, oh no, that's why they do it exactly like that. That's like why, right? It's the same thing with people who are like, oh, this is how we used to do it at this other company or I worked at or whatever. It's like, Okay, but does it work here? Like, the answer may very well be yes, right? But we should have a good reason for those things, right? And I think a lot of people are kind of just like, oh cool, that's awesome. I see that like, you know, that thing is working really, well. Let's just like go in and, copy it, right? And I think that just like, we are in the very, early innings of all of this, like agent stuff. What is the form factor? What is the experience? What is whatever, right? And like, you know, in the early internet, like we, you know, it was DOS and then it was like browsers, and then it was like mobile, and it was like all these different form factors that you couldn't possibly like comprehend as you were kind of going and, doing that. So if you think that you have found the like perfect form factor or anything else like that, like, and we also believe that we're in the early innings, like you're just wrong. Like there's so much more experimentation to go in and, do. So like try and go and experiment, right? And you may just end up with the same conventional answer that somebody else ended up, back, but you know, is the, other side, which is Somebody else—

Nakul Mandan:

you don't need to reinvent the wheel on everything. You're already reinventing the wheel in terms of infrastructure for coding agents, or a coding agent world. Do you really need to reinvent the wheel on how enterprise sales is done or other aspects of the business can be run?

Jake Cooper:

Yeah, so I, I mean, it's like the, classic like build versus buy philosophy, right? Like, so I think you should, you should really pick your battles. I think the folks over at Tailscale have this like wonderful concept that they use called innovation tokens where there's like, where are we gonna use our innovation tokens, right? Because if you're just like, I'm gonna build on everything from scratch, I'm gonna do all this other— you're just gonna end up with this like mountain of things that like aren't yet completed, or like, you know, maybe you are kind of reinventing the wheel or anything else like that. and so the moment you realize you've kind of like made that kind of first full circle, like that's, a point for you to stop and be like, okay, well, all right, great. Like, let's examine what we've got off the shelf and what we can kind of go in and fit in here, you know.

Nakul Mandan:

Do you also then gut check— let's say you arrive at a solution on some, business problem internally, do you then also do gut check that, hey, this is the best way based— like, is there a, some, market check on how others are doing in a way that could be better than what you arrived at as a conclusion?

Jake Cooper:

No, we, check with our users. Like we, built tooling internally to go in and talk directly with our users, to go in and get feedback. Everything is like layered into this thing that we call Central Station, which will automatically go in and aggregate sentiment, feedback, feature launches, all of this other thing. We built a whole like CMS on, in, this machine as well to go in and do launches. We gotta check with the users, we gotta check with ourselves. We want to use like the product and make sure that the product is really, good and that we're like proud of it and we're like pushing something out that we're like Great. We would use this. I will use this every day, right? Because we dogfood a bunch internally, not enough right now. And we're pushing a lot more on like, hey, how do we go in and use all of these like cloud agent primitives, all this other stuff? Like nobody should be running anything on your, on your machine, right? If we cannot model this thing on this like infinite virtual computer in the sky, we have to figure out a way to go in and do it. And I think dogfooding is just like one of the most important, impactful things that you can go in and do, right? And there was a period where we just weren't doing enough of it, right? And you can just see it in the product, like people who use their own product, right?

Nakul Mandan:

So user feedback makes sense on the product side, but what about pricing? What about sales-related or go-to-market-related things? Like, do you at all feel like, hey, if somebody like a Notion or Cursor or somebody else has solved a pricing and packaging problem, you want that as a final check on your solutions?

Jake Cooper:

Or— yeah, so I'm not gonna say that like we're just going out here and like putting our head in the sand and being like, hey, like, you know, like we know what we're doing, all this other stuff, right? Like we have, you know, channels internally like market intel and water cooler and stuff like that, right? Like recently, like turbopuffer did a really, great like ad with the cow. I don't know if you saw it. Or whatever, but we're like, oh, okay, cool, that's really, good, right? And so I think for things like, you know, marketing, whatever, like we go in and, say like, which companies are inspiring us to go in and do these things, right? Like, and what about those things are, kind of like inspirational, right? you know, I think Ramp is on that list. They do an exceptional job at a lot of these things, right? And then we go and we try and reverse engineer like, what is the way that we want to go in and do it, right? We say internally a lot like, what is the rich taste we want to go in and, do into this thing, right? Like, why are we doing this slightly a little bit different, right? Like, okay, for if, most people are doing X, how can we do just like something that's like Just like we'll stand out a little bit more, right?

Nakul Mandan:

There's another aspect where you're slightly different, not unique, but slightly different than a lot of the conventional wisdom is you're a solo founder and you have a very unique perspective there. So tell me, did you choose to be a solo founder or did you go looking for somebody and ended up not finding somebody that you wanted?

Jake Cooper:

I've never like done the like co-founder dating kind of thing or whatever. What's the Carl Jung quote? It's like, it's like an alchemist, and he's like giving advice to like one of his peers and it's like, if you do your work in earnest, unknown friends will come and find you along the way. Right. And I like love that. Love that quote because, you know, we just kind of have like continued to just move forward, right? Like our founding engineer joined us like probably a month after starting the company, right? And like always just wanted to make sure that we could hire wonderful people, do better but fewer folks, to go in and, move more things forward, right? Because then if you have those people, then you're not gonna get stuck into this thing where it's like, oh, you have hundreds of people and like nobody can make a decision, right? Like we work on a lot on DRIs internally, right? So somebody is going to, gonna make a call. I will purposely put myself into meetings with like 3 people so that I can be outvoted, right? Sometimes I am, I am wrong. Sometimes I am right, it doesn't really matter, right? Because so long as you're able to go on and move it forward, like, Notion gives you information, right?

Nakul Mandan:

But was the solo founder thing a, a choice, a very intentional choice?

Jake Cooper:

You didn't know? No, it was just like, I didn't, like, I didn't really start this as a company, right? Like, I was like working on this like prototype. I had like just left Uber at that, at that point. And I was like, well, like, how do you, like, you know, I'm gonna need databases for stuff, right? Like, let me just set up this like crappy box with databases on it. And like, I'll come back to it as I'm like working on like things that I'm gonna I'm gonna go start myself And then like, I was just kind of working on it and I kept coming back to it. and I was like, hey, I built this thing, you know, and then I kind of put it out on Twitter, at the time I was like, hey, I built this thing, like, who should I go in and talk to? Right. and I was like, oh, I'm looking for a job, right? Like, I wanna work on this, thing. Like, you know, I'll, move over here. But people had kind of taken it more as like, oh, maybe you're raising money or whatever. Cuz this is like, you know, 2020 or whatever. So it was like, there's money flowing around. And so like, I'd never— it, was never like a fully intentional thing until it got to a point where like, I had no job offers, but, few angel checks, and I was like, all right, screw it, like, I, I can always go back, you know, so just kind of like rolled through that. And then like shortly after, you know, like founding engineer joined us and then like a couple more people and stuff like that. So it just kind of like rolled that way. It wasn't like an intentional thing, you know.

Nakul Mandan:

You know, when we were talking, you also had this point of view that, hey, one could argue that solo founder should be the default because it's easier to handle all kinds of decision-making and disagreements in your own head versus with two different people. Could you talk more about that? You said something like, I think two is the worst number of co-founders.

Jake Cooper:

Yeah, well, two, is, two is awful. Two is awful. Like, it's like, it's like, do they make couples counselors for co-founders? Like, no, like you always have a disagreement, right? Like, how do you, how do you, know, deal with that? I guess like, you know, the, conventionalism is you just split down your zone of excellence, right? Because you're like, oh, this is my zone. It's like, sales, marketing, go-to-market, whatever. This is, this is the, you know, your zone, which is, engineering, product design, or anything else like that. But yeah, I think that like, you know, having, two is like, you're just consistently like, okay, well, Well, you know, you need a third, person to like go in and make that work, right? And like me, I'm consistently pulling in input, right? Like I'm, I work like sunup to sundown, Monday to Friday, and then I try and like, you know, close my laptop until like Sunday afternoon, do a little like full 36 hours of just nothing. my brain will just like, it doesn't really shut off work, right? And so it's like just analog time, writing, thinking, you know, maybe go on a hike or something like that. But yeah, I, I just don't know how people like deal with that. Cause I'm pulling in so much input all the time. I'm thinking about all these things. I'm making decisions. I'm like, I'm like, should we do this? Should we do that? Right? Like, it's, difficult to even like get consensus within myself versus like, you know, it's like, oh, I gotta go to this person and like, maybe they're not in the office. I gotta go find them. Right? It's so, slow.

Nakul Mandan:

Do you miss having a co-founder in certain aspects of running a business today?

Jake Cooper:

I think it's probably just like, you know, I, I've never done that, so I, I can't know. I think there's like, with everything, it's benefits and trade-offs. You're like trading something off, right? and so like the interesting thing about being a solo founder is like, I am the last and only person that picks up the phone, right? Like if somebody has a problem for me, like it means that like probably the whole rest of the org could not solve that problem, right? And if I don't pick up that phone or solve that like problem today, I have that problem plus every single other problem that I had tomorrow, right? And so, yeah, I think it's, just like if theoretically one could part and parcel out, okay, cool, you're just going to do sales, go-to-market, all those other things. And I never had to go in and think about some of those things. And I, I didn't like them. Like I actually I like many aspects about, those things. I think that would be a really great deal. I think it would be amazing to just like rid yourself of that thing. I happen to like really enjoy tinkering. I happen to really enjoy every single part of building a business. And so for me it's like, oh, problem that nobody else can kind of go and solve and bubbles up the, marketing stack or bubbles up the infrastructure stack or bubbles up wherever. It's kind of just like, oh, there's a new, there's a new problem every day, you know? Right.

Nakul Mandan:

Do you feel like the solo founder thing also works better? Between a certain— for a certain phase of the company, 10 people to 100 people, let's say, and at some point it does— you do want multiple leaders at the very, top who have the cultural relevance in the company and have the cultural context of the company to scale. Like, do you feel there's a phase where a solo founder is better versus 2 founders and 3 founders, or do you feel like—

Jake Cooper:

I'm not sure, like, it's one of those things where, like, I'm not yet there. Like, we're still building out a bunch of the— because we're still like 40 people internally, right? so we still have like some sections that have like, you know, leaders internally.

Nakul Mandan:

You've built a pretty lean team, 40 people, near $100 million of ARR, or maybe right around that.

Jake Cooper:

Pretty close, just shy, but we'll, be there. It depends when this rolls out.

Nakul Mandan:

Yeah, that's how close you are. So was that intention from day zero? Like, you know, as you— as this went from a project to a company, you were like, I'm going to go extremely lean, or—

Jake Cooper:

Yeah, I mean, for me, yes. like, I, I want to work with great people who like want to be great, or, good people who want to be great, right? Like good humans who like wanna go in and push themselves further, right? and like for me, I've always really, loved work. Like I, I like am, a workaholic. I have to pull myself away from a lot of these things. The kind of like goal would be to just find, you know, fewer folks who are just like more kind of work, aligned and they really love what they, go into and they wanna push themselves and they want somebody around who can kind of coach and like spot for them, right? And like let them like go for gold, right? And if they like, you know, miss something, you got people who are there, they're like kind of like a spotter with like a weightlifter or anything else like that. And so I think that was kind of the goal, right? I think— do you know Renaissance Tech? Yeah, the hedge fund or whatever. Very, you know, elite group of folks, doing high leverage stuff. you know, and, just like they're really, excited about the work that they are going and doing, and they just like do it at like a world-class level.

Nakul Mandan:

Correct me if I'm wrong, when you have this lean mindset, people who are going hard at it, you're not saying necessarily just work 997 or whatever, right? You're not saying that. You're just saying excellent people, but like, what goes into that operating philosophy? How do you get every person that much more leverage? Is it systems building? Like, can you do go deeper? Because you are in the— you might not appreciate it how unique you are. 40 people and $100 million of ARR is extremely fucking unique.

Jake Cooper:

Yeah, it's, super, unique and impressive. Yeah, I appreciate that. it is a lot of work. Yeah, I mean, it's, a lot of like systems building, right? it's a lot of like saying We, say no to like 90% of stuff internally, right? Where we're just like, no, like as of right now, we're not going to conferences or like as of right now, we're not doing like this specific stuff or as of right now, like this part of the product we're not pursuing, right? Like we didn't pursue a bunch of the AI stuff until like probably 2023-ish or whatever. And it took a while, right? you know, cause we say no, we say no, we say no. And then when it's inevitable, we say yes. Like that's what the thing we're going to go in and, do, right? So it's a lot of systems building. It's a lot of kind of like argue on like our values and, like I think it took me like 3 or 4 years to like come up with like the values that we had. And like when, we started the company, I was like, oh, the values are important. So I like written them down and like 6 months later they're completely wrong or whatever. And so it's like a, just a lot of work, right? To like go in and, get there, you know?

Nakul Mandan:

When you say no to 90%, what makes the cut? What makes the cut? What's the hurdle that something has to clear for it to say, yeah, we are doing this, we are choosing to do this.

Jake Cooper:

Yeah, I think it has to be like almost the right time. I think like, you know, our founding marketer has this, saying that I really like internally, which is, disagreements of timing are my favorite to have because all disagreements are disagreements of timing. And so if it's like, oh, why is it the right time for us to go in and do this? And so the way we do stuff internally is we like write a small abstract, right? And, the, last line in the abstract, right? you know, you got background, a little problem, you know, what problem or opportunity is there? The last line just says, why now? Like, why are we doing this now? Right? Like, you know, can it come slightly later? Do we need to go in and do it now? Like all of those other things, right? And then, you know, these things flow through like one channel where everybody kind of watches them, et cetera. And then we plan quarterly into like going to roll them out. And those things end up, you know, like we talk about them, we spend an entire week like doing this planning thing, which is like crazy for like a, you know, early stage startup or like whatever. But like we want to talk and we want to like, you know, scrutinize this thing and be like, hey, listen, should we do this now or should we do this later? You know, and, so we just end up the default response to like say for a lot of these things is no, like we're just not going to go in and do that, right? But then you hear the same thing again and again and again, right? And like, great ideas are inevitable. You just can't escape them or whatever, right? And so they will just keep coming up, right? And then after a while, it's like, we, have to do this now. We just have to do it, right? Somebody gets fed up and they're like, we're doing this, right? It's like, great, awesome.

Nakul Mandan:

So everything you just said, being very intentional on the initiatives you take, timing, everything, but is it also that you try to solve things more with systems than people? Because, you know, like, we, work with so many companies and founders feel like, hey, if, we want to go, down this initiative, we need this many people. Like you have one sales rep, one product marketer, and two solutions engineers, right? That—

Jake Cooper:

and, a growth marketer and then two DevRel folks.

Nakul Mandan:

Okay. So that's like a 7, 8 people go-to-market team.

Jake Cooper:

Totally. 100%.

Nakul Mandan:

The initiative is right, as in sales marketing needs to happen now at the company, but still it's lean. So, and it's very lean. So is it a function of also like systems building that you've kind of incorporated at every aspect of the business, or—

Jake Cooper:

So I would say that like we don't want to do things manually. And so if we're doing something right, it should be like, grit, process, repeat, right? So start, by doing it manually. You have— you actually have to do it manually initially just to like really understand what this thing is, right? And even to this day, for new roles, like we will ask that folks like go in and do this thing manually so that we know what we're going to go in and hire for, right? We know what the jobs to be done are that like fit directly into that JD, and then we will go in source specifically for those folks. And that's how we find exactly the right person for a lot of this stuff.

Nakul Mandan:

Is that part of your recruiting criteria that the person has to be a systems thinker? Because often when execs join, and I don't know how big is your exec team, their mindset from prior companies is they solve problems by hiring their people.

Jake Cooper:

Yeah, it's, an adjustment definitely to go in and work like that, right? So we've, been very, systems forward and we have, recruiting writing, code and automations and systems. We have sales writing code and automation and systems, right? Like We wanna be able to make sure that folks can go in and build these systems because otherwise there's no way in which you can build this amount of leverage.

Nakul Mandan:

So, okay, so let's shift gears towards recruiting. Maybe start with execs and then we— I'd love to go into even how you recruit IC levels, but when you're recruiting execs, what are you looking for?

Jake Cooper:

So, execs are, odd in the sense that like we haven't really recruited too many of them, but we usually look for people who are like gritty and will like go in and do this kind of like themselves. They really, want you know, to get after the craft in the zone versus actually just like almost, you know, you see this at big companies, you see that like whole empire building of like how many, how, can I hire all these people? Or like, how can I bring all of my friends from, you know, prior company or anything else like that? Like those modalities don't tend to work like super well for us because we want people who, you know, they are, they're closer to like player coaches, right? It's like, oh, you can like, you can play ball, right? Like our head of engineering is out here being like, hey, I can't parallelize enough agents to like get through all the shit that I'm like trying to go and, get. And we're like great, awesome. Like, that's great. Right? and so we just look for folks who are going to be able to almost like go and do some category of the work and then like really give people like core feedback, figure out what's slowing them down, like empower them to like make decisions and all of those other things. Right? And there's like a whole like series of, you know, criteria that we're, looking for internally as, part of that. But like drive and autonomy and like, you know, do they, do they, do they fit the values, you know, and the purpose, why are they here?

Nakul Mandan:

Right?

Jake Cooper:

Why do they need to be at Railway at this specific time? right? Like all of those things are like very, core and important.

Nakul Mandan:

But have you been able to codify how to look for those traits of that system scaling mindset, that, people will ask in an interview? Like, are there specific questions you go, to, to test whether this is a good cultural fit for the lean mindset that you guys have?

Jake Cooper:

No, not really, other than the, obvious one of like, can they build with leverage systems, stuff like that, you know, right? Like those are things that we, like are like, you know, looking for.

Nakul Mandan:

Can you talk about, without naming the person, like a hire that did not work out and what have you corrected in your own recruiting philosophy over the years?

Jake Cooper:

So at one point, really early on in the, company, we didn't run background checks, like just to— like we're really like early stage startup or whatever, right? And then we hired this person and, pretty quickly it like didn't work out for other reasons or whatever, right? And then we kind of found out that there was like some stuff that like should have come up in a background check if we, we'd done that. And we're like, yeah, so now we do background checks, I think is probably the, the, one that like stands out to me in terms of, you know, putting yourself out there, learning, you know, stuff like that.

Nakul Mandan:

But in terms of like recruiting philosophy or management, like, yeah, like that one feels like a process check that you should have. But is, there some trait or something that you look for now that you were missing earlier on or vice versa?

Jake Cooper:

I don't, I don't think it's a trait that we were missing earlier on. It's just like, it just continues to become like really, important, which is like, is this person a strong swimmer? Can they operate autonomously? Are they going to be able to move the business forward, change the trajectory of the company, things like that? Like, that's what we're looking for in every single person that we bring on board, right? So, I would say that one's like just a really important one.

Nakul Mandan:

And then last night when we were talking about recruiting, we also talked about how comp, has moved up and taken away the— you know, you're competing with Labs and some of the most well-funded companies ever. Uber. So how have you thought about comp for a company of your scale? And yeah, how do you compete for talent in today's world?

Jake Cooper:

Yeah, so I think the, like, remote thing definitely helps us, right, in terms of, like, being able to get talent across the, globe. you know, somebody else at the dinner last night was like, yeah, the, Bay Area talent area, market is kind of cooked because, like, from the labs. I don't necessarily disagree, right, but I think that, like, great people with absurd amount of leverage— and agents, whatever they are, today, like, constitute leverage— should continue to see their like salaries go up and up and up. And I think like, you know, as, we go in and see more and more and more levels of like production, it's, you know, it's, completely absurd that like somebody like Andrew, you know, moving from like Meta, like Anthropic with like a $1.5 billion pay package, you know, like it's absurd, right? But so too was acquiring Instagram for a billion dollars or whatever, right? Like it still ends up being a net positive ultimately there. And you're just seeing this like enormous, like there's no ceiling for like how good somebody can possibly be with these machines, especially right now because the frontier is so, jagged and you have this diffusion of like some people are running 50 agents and creating, you know, the worst thing ever that will just destroy your entire codebase. And then some people are running 50 agents and they are creating the equivalent of the Sistine Chapel, which just actually you should just bulldoze your entire codebase and replace it kind of with, that, right? And so it's just And it's, tough to tell sometimes too with the whole thing. So I just think that like our approach for talent is like never be stingy with, talent, right? And just look like, you know, I'm evaluating comp like, you know, every half, it's probably moving every quarter, right? Where it's just like, are we making sure that these folks are in the best possible situation, right? Like from both an equity ownership perspective as well as like—

Nakul Mandan:

Even for your own employees, existing employees, your—

Jake Cooper:

Oh yeah, Especially like, and I think people forget that like all the time, right? Like you need to be able to make sure that like people are being retained. People are excited about the work that they're doing, all of these other things, right? Like, I'm a big, believer in like servant leadership, right? Like, and, that's what we see in player coaches too, right? Like where they're just like, all right, here's this, you know, pardon my French, dogshit work that I'm just going to go in and grab, right? And just make sure that you guys are able to go in and do those things. And then also set the expectation that like you're going to do those things really, well, right? We're going to talk about your work. We're going to push it in public. We're going to go and make sure to chase down users to be like, what is the feedback? Tell us the feedback, you know? And just making sure that environment is not just like excellent [inaudible] from like a, you know, comp, whatever, but also fulfillment, right? Also they're working on interesting problems, right? Like part of the reason to start the company and like, you know, calculus that goes on in my brain early on of like, do I want to run this company or do I not want to run this company is like, there is so much interesting stuff, like so much interesting stuff, right? Like we're doing like crazy bare metal networking stuff. We're doing like on the go-to-market side, we're figuring out how to like incept the models to like, recommend us more and more and more. And we're figuring out how to go in and teach the models to essentially put its like, you know, hand or tendril or whatever you want to kind of, on, the infrastructure that we've got, right? Like they're all such interesting problems, right? So I think you just need to make sure that like, you know, in today's age, right? Especially like, you know, if we get to a point where like maybe AI is so good that people have to work less, like if you want to be productive and you want to move, you know, society forward and have purpose, like you have to be really, good at all of those things.

Nakul Mandan:

Has AI changed how you hire engineers in terms of the process itself? Are you— applying a different process to the point you were making, like somebody could spin up 50 agents and destroy your codebase. Somebody could do the exact opposite. How are you testing for this?

Jake Cooper:

Yeah, so I think there was a, there was a step function where actually, you know, we, almost blinked after a bunch of the AI stuff, you know, it was like really, good and we realized we hadn't updated like our, technical interview.

Nakul Mandan:

Yeah. And when was that actually just for—

Jake Cooper:

so that was like, you know, we like, there was, there was a couple moments, right? There was like moments in like 2024 where we're like, okay, like you can definitely get mileage out of this thing. And then there's like kind of like, you know, Opus 4.5 moment, which I think is like the one that most people are familiar with where it just like hits everybody like a ton of bricks and you're like, oh, this is crazy, right? And so for us, that was like, you know, I had plans going into December about, okay, cool, we're gonna do enterprise sales and all of these other things. We just raised our round.

Nakul Mandan:

This was December of '25.

Jake Cooper:

Yeah, December of '25, right? And I was like, okay, well, we got this agent roadmap, we got this enterprise roadmap. And I, I go to like, you know, my— we do 2 weeks off at, winter, break. I come back from winter break and I'm like, yeah, the enterprise stuff is, gonna have to wait. you know, it's— we're, all in on, these things, like, you know, we just have to move a lot of this stuff forward. And so I think to answer your question, it changes the, ceiling. It changes the ambition. It like, as you, as you kind of go through these steps, right? And the new like OpenAI, model, Anthropic, I think it's so like, it's still another step forward and it's so good, right? And so you have to consistently raise your bar for whatever the ambition is of the thing that you are doing, right? And so we look for that. We always do, you know, in the technical interview, we will say, Hey, bring whatever you want ahead of time. You can do it cold in the interview if you want or whatever. If you're not like, I'm not gonna prepare, you can spend an absurd amount of time preparing for it. Doesn't really matter to us. And we just like go through that prototype, that thing, that piece of work, et cetera, with you. And so I think we just have to consistently almost be like, oh man, the models are so good. Like what you can do in some period of time, we gotta like move it further and further, right? And then you have to actually like evaluate people on things inside their work, right? Not just the, oh my God, that's impressive or whatever, but, and this was always the case, right? Like, are they using the right patterns? Are they like doing the right things? Like, all of this kind of other stuff, right? So it's just a deep amount of scrutiny.

Nakul Mandan:

Has accordingly all of this exploded your roadmap, and how are you contain— constraining it to still be the 90/10 of product? Let's talk about that. Like, what do you say no to on product today when everything can be built so fast?

Jake Cooper:

Incidentally, I believe that there's almost like only 3 things left for you to do, you know, just as like a human or a builder or whatever. It's a— it's taste, structure, and agency Those are the only, the only two things, right? And, because those things are basically, or because those things are hard and everything else is basically free now with a lot of the agents, it does come down a lot to like, what do we build and how do we build it and how do we present it to the user and how do we go and give it to those folks, right? And so I think for us, that's why, you know, we do spend a lot of that time on like discussing those ideas, right? Because once the roadmap is locked, right, you're just like, okay, great, that's my swim lane. I'm just gonna like, burn it down and then expect everybody else to kind of like burn down their, section of things, right? and so that's kind of how we think about it is like, you know, lots of, you know, I, I believe in like cycles of like expansion and then compression. And so we do a lot of idea expansion and then we do, you know, a lot of implementation compression.

Nakul Mandan:

Going back on the people side, for, years, I think you had no onboarding for people.

Jake Cooper:

People would start and it was like, it was like, hey, welcome to Railway, grab a shovel, you know?

Nakul Mandan:

Yeah. And now you've shifted. Tell us why, that did not work and what, made you feel like, hey, you know, we need some kind of an onboarding. And then I think you have a pretty unique onboarding even now. So we'll dive into that 6-week onboarding thing that you have now.

Jake Cooper:

The success of, any individual at the organization is measured by, like, can you get that person to do the best work of their life and how quickly can you get them to do the best work of their life? And so I think, originally it was, like, easy because the codebase was small and the context was, like, pretty, like, whatever. And then over time, like, that context would, like, grow and grow and grow, right? and, like, there's various different sections of— I think, I think, I don't think we're unique in this sense, like, but I think every company has various sections of like some parts of your company are, more well-formed than others, right? Maybe like the basement isn't like renovated or whatever. And you're like, yeah, it's kind of like dicey, right? And then some parts actually just like, you're just like, oh, there's tech debt or there's landmines or whatever, right? Like you should probably not be in there. Setting people up with their own devices or whatever to be like, all right, figure it out or whatever. Like they just don't have that context, right? And there's like, in my mind, I don't wanna say there's no amount of excellence because any amount of excellence you can figure it out, but we were basically like, well, like, why are people— why are excellent people using their time? Like, if we're saying we're gonna recruit and pull in these excellent people, why are they using it to like navigate this thing that like will take us literally, I don't know, an hour and a half on like a Sunday to like put together a really, structured, awesome 6 to, you know, 8-week onboarding plan that we can go and sit down with them based on their background, based where we are in a company, right? If we wanna go in and hire fewer, more amazing, folks, like how do we set them up for as much success at, out of the gate? And then Pushmore is excellent.

Nakul Mandan:

But the thing that, was interesting is that even when there was no onboarding, you were like, you'd let the person who's joining figure out what the company needs and then run with it. Now it's the third part of the onboarding. So that was the interesting thing where it wasn't just pick a shovel and do. It's more like you were letting the person decide what does the company need, you figure it out, and then go and solve that, right? That was the initial onboarding.

Jake Cooper:

It's just, it's just 3 things.

Nakul Mandan:

And now it's 3 things.

Jake Cooper:

Yeah, it's— I'm a big, like, 3 things person, just broadly speaking. trust, clarity, judgment. That's like 2 weeks of trust, 2 weeks of clarity, 2 weeks of judgment.

Nakul Mandan:

What does 2 weeks of trust and 2 weeks of clarity mean in onboarding?

Jake Cooper:

So, these are tasks, small tickets or anything else like that. They're really straightforward.

Nakul Mandan:

Like you give them the task to solve.

Jake Cooper:

We just basically like, hey, here's 5 things roughly around the, your zone of excellence that you want to go in and move towards. If you join us on the infrastructure side and you say you want to work on networking, we will set you up to go in and build networking. We have had people join and then, you know, within 9, 12 months, et cetera, they were owning the entire edge network, right? so those like 5 things to just start. Like boom, you can just crank those out, right? And we were just like, great, these should be easy. 2 weeks worth of whatever, right? Allows you to like get your sea legs, like talk to people.

Nakul Mandan:

That's them earning your trust as a company.

Jake Cooper:

Yeah, it was like, it's like seeing, it's like seeing those reps kind of close, right? Where everybody's like, great, this person's moving. I can see what they're working on. I can see how they're working. I can see all these other things. Oh, amazing. I'm so excited that somebody is like going and, joining us to work on, trust and safety. Like I've had so many ideas. Is, that's awesome, I can go in and grab time with this person, right? And we say, okay, cool, like grab some time with some folks, right? And then internally we also say, you know, we have an onboarding channel and the, onboarding channel gets broadcast to basically the entire company. So everybody can kind of see, we, try and help people out as they're kind of joining, right? And it gets trust on both sides, right? Because like just joining a new job is never— it's just never easy. And you should just acknowledge that, right? And to be like, oh, okay, cool, you should feel good, you should get reps, you should feel awesome and comfortable. And then you kind of move on to these like, what we call problems, right? Which is instead of a task, which you're like, oh, okay, it's like pretty straightforward to go in and do this. A problem is like there's multiple solutions, right? You'll, have to figure something, out, right? set up some trade-offs, some benefits, like, all of those other things, right? And so we have people write internally a request for comment, which is just like a little bit of background. What's the problem? What am I gonna go in and do? And then like, what's, what are alternative solutions to go in and do this? And then we can go and talk with this person, right? And we can say, okay, well, like, actually that's like probably like, you know, that looks like 8 weeks worth of work, right? Like, here's how we can trim that down to like just the stuff that we need, because again, we're a big, let's do the 10%, maybe 20% that matters, right? And the other stuff we can, we can bring later, right? And incidentally, that's actually a really, good thing for onboardings, because if you leave those 80%, it's actually very, clear the jobs to be done when somebody joins your organization. It's very, clear how you can build an onboarding plan when most of the P2 requirements just have not been completed, right? and then so then we move folks, you know, from the, you know, the, trust to the, clarity and then to the judgment, which is what do you think the business needs at this point in time? That is the only prompt. That's it, right? And then you're basically like, okay, well, like, we should be doing this. Our observability story is, you know, dog water, and we need to go and improve these things. Like, great, we think that that's correct, right? And we also kind of, you know, we, will give this, kind of like structure ahead of time, and we, you know, tell people to use that thing internally called a scratchpad. We just use Notion, and you just like dump ideas into this. And so by the time you end up with this thing, you have a bunch of, you know, things that either people have been like, oh, you brought in to work on trust and safety again. I had these kind of thoughts or whatever. And then you kind of have these like cards that you can kind of go through and be like, oh yeah, I want to do like, I'm gonna do this and this, right? Or I'm gonna do just this one thing cuz it's like gonna take me probably like 3 or 4 weeks. Boom, right? And what that sets people up to do is they basically get, you know, in university back in Canada, you would have almost like a capstone project, right? Where you just like put something together, right? And it was like impressive and you like build something from scratch, all this other stuff. You do the same thing basically, internally.

Nakul Mandan:

Does each person then eventually end up having their own zone? You, call it zone, right? The agency. Amazing thing. So is it like 40 people all running on projects that they chose that the company needs?

Jake Cooper:

No, because that would be a total tire fire. Yeah, because there is a bunch of stuff that just has to go in and get done, right? you would just have disproportionate amounts of people working on all the hype, hot AI, whatever kind of things. we try and make sure that folks are set up to, take shots unilaterally as part of onboarding because we want those folks to, be like, drag us to the new frontier. You know, you're new, you're new, and there's a specific time in which, you know, you can join an organization and, not be brainwashed, right? Because then after a certain point, you're just like, you know, we— the culture, these are the norms, all of that other stuff, right? And so we want people to kind of drag us towards those things. But, you know, that kind of pulls back to like the planning phase where we're like, okay, here's the things that have to get done this quarter, right? And maybe you're just like, you know, on the product side, you're like, oh man, you know, like I want to work on these things. And infrastructure is like, hey man, like we still have problems with the crypto miners, you know, and you're, letting these people in the front door or whatever, right? We need this kind of work to be done, right? And then we go into the product zone. Say, well, listen, 70% of this stuff is steak, 30% of it is vegetables. Let's break it down. We're each going to have a vertical. We're going to go in and do that. We're all going to sign our names on the things that we're going to go and get done. And by the end of the quarter, we're going to go and complete those things, right? Hell or high water, we're going to go and complete those things, right? and so it's, a blend between like having people go in and drive those things and also, you know, again, jobs to be done.

Nakul Mandan:

Let's shift towards the overall— at the company level, you're remote as a culture.

Jake Cooper:

Was that Was that a choice or was that a happenstance because you started during COVID It was both basically, because I had worked originally at Wolfram Research, which was all remote. and yeah, it was just, it was one of those things, where I was like, I enjoyed the culture. It allowed me to like be very heads down on, stuff. I always felt like I could move the most important thing of the week forward. And so, you know, we, started the, you know, I started the company and, then, I was like, oh, it's just gonna be remote basically. Because our founding engineer was in Montreal at the time, right? and it like never even really crossed my mind to like have this in, person because remote allows us to pull the best people from literally every, facet of the world, right? Like I have so many fond memories of this. We've got this wonderful, guy, on our team, who works in Thailand. he goes by the name Five, like the, number 5. I have so many fond memories of like working directly with Five where like I, I, I will like go to sleep and be like, hey, Five, can you work on this thing? Like I'll hand this like part off to you, right? And he's like, yep, cool, great. And I'll wake up and it's done. And it's the closest thing you can get to time travel. It's crazy. And they can do the same thing on their side. Be like, okay, let me pass this thing back to you, right? You know? And so it's like, to me, it was like never, even a consideration.

Nakul Mandan:

Is there a downside that you found, like managing people remotely? Have you had to instill systems and culture?

Jake Cooper:

There is enormous downside. yeah, you have to build, not only the systems to go in and support your organization, to go in and make this thing work remotely in whatever kind of way you want, which thankfully because of AI, you can move a lot faster to go into. And then you have to be competitive in your core product, right? Like, and so you have to build these tools or you have to buy these tools or you have to go and pull them off people. Like, distilling information remotely is more difficult, right? We used to like do almost exclusively written stuff, very, few meetings. Now we've added like, you know, transcriber bots that will join the meetings and like automatically attach, the context to the like right, like planning item, right? it's built on top of, you know, the principles from Toyota Production Systems, which is the original kind of Kanban before it got bastardized and brought to the West. And so like, those are things that we've had to go and build and, set up, right? and so yeah, 100%. And then I think also like less talked about is that it's kind of the whole like you need strong swimmers, kind of thing. Like we've, tried to hire more junior folks and it's like one of those things where you're just like, maybe this time it'll work and it, doesn't work, right? You just need folks who are going to be able to kind of like turn those things over, right? And there are exceptions to that, rule, you know, and you can go in and find folks and we're getting better and better and better at that. But it's, just trade-offs. Everything is, everything is trade-offs, you know?

Nakul Mandan:

So, and then your values, they spell HARD. So, honor, acceleration, reliability, directness. How did those come about?

Jake Cooper:

Oh, that was like, that was like 3 or 4 years and like hires that didn't work out and like, you know, stuff where I was just like, I'm trying to teach people these things that like, they matter a lot in terms of, you know, like make honorable decisions, right? We don't want to be the company that's like with dark patterns everywhere and this stuff, right? we want to like be accelerating. Like this is why we've shipped on Fridays and we've shipped a changelog. For every week on Fridays since like 2021, right? You, wanna be reliable, right? Like you don't wanna be the coworker that's like, oh, I got paged and it's like, it is what it is, man. Right? It's like, no, you, wanna be reliable. And then you wanna be direct too, cuz like people waste, so much time like holding in feedback, emotion, like whatever, all of those other things. Like if you're direct and you just like cut to the 10% of the, problem that matters, if you're just basically like, hey, listen, you know, feeling these things that we need to move this part forward, why is it not moving forward? Let's talk about those things. Right? And the nice thing about the, values in, my mind is that like they're complementary, if that makes sense, right? and so let's say you were gonna go in and give feedback to somebody, right? Like you'd be like, hey, like, you, suck and here's why, right? Like it's very direct, right? Is it honorable? Like even close or whatever, right? Absolutely not, right? And so you have to kind of move with each one of these values, right? You know, if you're, you know, okay, you're working on feedback forever or whatever, right? And then you like don't go in and, give it, you know, it's like, is it accelerating, right? And so each one of these things is a balancing engine, right? And when you hit those things like perfectly, it allows you to kind of like move a lot of this stuff forward, right?

Nakul Mandan:

Which one is the hardest to live up to? Like as an example, has honor cost you something?

Jake Cooper:

Yeah, it's— I was gonna say it's, honor, I would say for sure. It's really, hard, to go in and do that. Like, right, there are things that like other people are doing. There are things that like, you know, like I think we probably could have like probably could have raised more money if we wanted to like oversell like the vision for all of, all of these other things, right? And then it's like a delicate balance too. Because you, want to go in and sell the vision, all this other stuff, right? But I think honor is probably the one where it's cost, you know, the most, you know.

Nakul Mandan:

And an example of that is like you don't oversell.

Jake Cooper:

Yeah, you just don't oversell. Yeah, right. Like we're just like, you know, we, don't want to put ourselves into a spot where we're like, we're saying, yeah, the product is going to do all these things. Like it's going to, you know, it's going to wash your car, it's going to do all this, all this other stuff, right? Like just trust us, like all of a sudden, and then you violate that trust, right? Like, you know, it's, just not a way that we want to kind of like conduct business, right? And that goes to, you know, how we purport our brand. It goes down to like like, you know, how we call incidents, right? Like we, you know, we had some reliability issues earlier in the year, right? And we're calling every one of these issues. We're saying, here's what's happened. You know, we turn around the, incident response in, like, you know, less than 24 hours. Earlier in the year when we were growing, like, when that step function had kind of changed and we were like, oh, hey, listen, like some of our engineering things, you know, normally in engineering you build essentially for about 18 months and you start to replace it around a year, right? Like because of the like jagged frontier and the lurch forward, we had some reliability issues, right?. And, you know, we're out here, we're calling every single incident, we're putting it out on Twitter, we're going to say, hey, here's what happened, you know, within 24 hours, like every single step like goes out, right? Because, you know, you can go and you can, you know, pull away from that, but then, you know, you're leaving your customers hanging, right? And so you have to kind of like take it on the chin, eat that, and then, and then, you know, you can't avoid it from a company perspective. It's like reliability has to be the number one thing. We have to go in and do it. It's a value, it's a thing. Like there's no way to go in and avoid it, right? And so we went like heads down and drilled all those things out, right? And now we're in such a better spot. And it was like pretty quick to like go and resolve a bunch of those things. So again, it's on directness, it's on reliability, it's on honor. And then you want acceleration. You want to do— you do it today, don't do it tomorrow. Like ship it, ship the thing today, right? You might break something, you might page somebody, right? Okay, what happens in that situation? Well, you know, you, rely on your coworkers. And then also you can maybe build a better system that says like, hey, when you merge a pull request, maybe you're automatically taking the pager for 30 minutes. That's the thing we do, right?

Nakul Mandan:

And one of your mantras is go hard, go home, which is go hard Monday to Friday and then communications, like all internal comms are muted till Sunday.

Jake Cooper:

Well, I mean, I can't do that for like company-wide, wide stuff essentially for, other folks. But like the way that I construct my week essentially is like on Sundays I'll do a little preparing in the evening. you know, I've got my to-do list on my, on my desk as I set things up for the week. And then Monday to Friday, like sunup to sundown, basically I am getting every single thing that I need to get done. And then, Friday evening, you know, at night my phone goes on DND, like do not disturb, right? between like the hours of Friday, evening to Sunday afternoon, basically, right? I try and like, I'll take my journal places, I'll like go and recover, I'll like go to the sauna, I'll like go in and do all these other things, right? And inevitably, because I care about business and all the work and all these other things, right? Like my brain will just naturally get pulled back to the, to these things, but I won't allow myself to like go back to a screen or I won't allow myself to do any of these things, right? And then if I need the space, I'll go in and recover, right? I encourage all of the folks internally, like, once in a while we are gonna have to break glass and we're gonna have to go in and work a weekend, right? Whether you are on call or anything else like that, or if we're just like, Jesus, we gotta get this thing done, this is the last mile, et cetera. That's a break glass function, right? It means that like our planning, our prioritization, we didn't say yes, no, whatever to the right stuff, and we have to make a temporary adjustment to go in and do this, right? But inevitably, like, if you're doing like 996 or anything else like that, you're getting what, like 20% more cycles? And like, that's really, at best, right? and like usually it's like, you know, probably closer to 5, maybe 10. If you use that, you know, decompression time deliberately, you can use it for strategy, you can use it for planning, you can use it for all of these other things, and you can make, you know, massive, step jumps, right? Like Jeff Bezos has that whole quote, about like in business once in a while, you know, in baseball or whatever, you can, you know, hit a home run and it's like 4 runs or, whatever, right? But in business there you can get a 1,000x home run, right? Like there are massive, step functions buried in inside of everybody's business that they can go and unlock by simply, like, what I would say, finding stillness. Because if you can find stillness, you can think about a lot of these things. You can go to the journal, you can be still with your mind, and like, it is crazy how well it works to simply just like slow down your brain. Stuff will just come out, like, it'll, just land itself on the paper where you're just like, oh, that's like the biggest— that's the biggest problem, right? Like, when you just like force your mind to be, calm and you just— oh, that was the biggest thing that we can move towards, this week to go and move the company forward. Great. That's the thing that ends up on my to-do list. That's the thing that's the highest priority for me.

Nakul Mandan:

You've said 2026 is a year when Railway plays on the world stage. What does that mean in terms of headcount and burn and spend? Because as the company then plays that, like, does it break from the traditional mold for Railway of being extremely lean?

Jake Cooper:

I think for us, we just like, you know, it's, a moment where we switch from almost like almost just surviving, if that makes sense, to like, how do we win these markets? Right? And, like, it's a, it's a shift for a startup to go in and make, right? Like for years and years and years, you're just like, how do we, how do we not run out of money? How do we like, you know, scrimp this, all this together, right? But at a certain point, like what got you here will not take you there, right? You have to make it like mentality switch. You have to basically say like, you know, like we've had conversations over the years with our, board and they're like enterprise sales, man. And I'm like, well, like we're hitting all these revenue targets, right? But like these are years where we're basically saying like, no, actually, like, you need enterprise penetration. You need an enterprise in sales. It's a people business. It's hard. It's hard as hell to build systems for a lot of those things that actually just kind of like pull down to, hey, these are personal relationships for folks, right? And so there are things that have to go in and change, and then there are things that we want to go in and keep the same, right? Like, how do we go and, you know, get to like, like 80 people next year, right? Like maybe 100. And then we're doing 500 mil, maybe more in like revenue, right? Like those would be the ideal goals, right? I think they're, ambitious, but, I think if you're not setting ambitious goals, then, you— especially right now, you're, not doing this correctly.

Nakul Mandan:

I would say 100 people and $500 million revenue is still like continuing on the current path in the sense that it, it's still very lean, extremely lean. Do you feel like there are principles of yours that got Railway here that you're going to divert away from? Like whether it's remote, whether it's lean, whether it's this first principles thinking, whether it's hiring more execs. Are there aspects of Railway that got you here that you are going to depart from?

Jake Cooper:

Yeah, I think, I think probably a, a big one, that I don't know if we'll, depart from, well, I think we, will and sort of already are, is like we were doing like, as few meetings as possible, just like try to do none amounts of meetings. The optimal number of meetings is zero. But I think with like the way that all of these things are moving really, quickly, you know, voice is a crazy, high bandwidth medium, right? You can get so much, so much, so much done on that. And, conversely, we don't use AI internally for any of the memos or things that we go in and put together, right? Like nobody should have to read the, thing that, you have slopped out, and just like pushed on, onto somebody else, right? And so to go in and put those things together, it takes a lot of time now, right? Like clarity is a lot of effort. Precision's a lot of effort, right? And now we've gotten really, good at almost pulling in context from all these meetings. And so you can just like pull people in and just like, you know, bang ideas back and forth, go and do all this other stuff, right? Literally have Claude agents listening to your voice call, kicking off jobs, putting these things together. What would it look like? All this other stuff, right? And then you can go in and take all of that expanded context now and compress it. And, so like, they're just different ways of working, I think, right? Where you just kind of take the old and then you like layer it in with the new, right? Like you, what you cannot, do is you cannot be a Luddite about about these things. You cannot ignore these things because like, again, they are just inevitable things that are going to go and occur, right? And so there'll be, I'm sure, plenty of those things that, will ultimately change. I think one strongly held belief is like, we should be remote, right? Like, you know, we want to make sure that like access is uniformly distributed, whether it is access to compute or whether it access as access to like ambition essentially, right? Like those things are really, important to us. It's just like core, consumers.

Nakul Mandan:

It's a good segue to the where the world is going, which I would love to get your take given you, have such a central vantage point on what's playing out in the world. So one of the things you've said is there's going to be 1000x more software in 5 years, maybe more actually, maybe even you're—

Jake Cooper:

I think it's— you're gonna have, you know, 100 times more people writing 100 times more software, so you have 10,000 times.

Nakul Mandan:

So what breaks in this world in the next 5 years? Like what breaks first as the world goes here?

Jake Cooper:

Yeah, I mean, I think just like compute generally breaks. I think you're already sort of seeing that where like I have friends who are just like, oh yeah, I'm running scraping jobs to find like available EC2 instances for like going and getting my stuff deployed, right? It's not just your one prepper friend or whatever. This is like many, friends who are just like, oh yeah, like we're running outta compute over here. So I think that that's like a thing that ultimately ends up breaking.

Nakul Mandan:

Do you think the world, the tech world has made enough investments that it'll be fine in 3 years? Or you think that we— everybody talks about the compute shortage is coming, but when does it actually hit?

Jake Cooper:

Well, we'll, have to go in and see. I guess what I'll say is I think people have invested a lot in the first order effects of AI, but have not invested a lot in the second order effects of AI, right? If Claude is free and it can— you can generate, you know, tons and tons, like we have so much tokens and so many, so much inference like coming online, what do you use tokens and inference to go in and generate code. has to run somewhere. I still think that we are undervaluing just how much that inference is going to come online, how fast these new models are, right? Like, you will get to a point, and you probably should, you know, arguably be now here now if you're on the frontier where you're not looking at most of these things, you're not even like looking and structuring the jobs that these agents are going and fanning off. You're basically just giving them carte blanche to come back with some sort of structure. And then you are beating into shape and you are setting up, you know, automations to go and receive events that say, okay, every single time you get a ticket or a customer support or a thing or whatever, see if you can fix it. If you can go in and do those things, right? And the amount of compute that you need to go in and do that is so, high. And so I think we're going to have like, we have an inference kind of crunch now, and then we'll broad strokes have a compute crunch, right? And like, I mean, what happens when Meta with like 3 billion daily active users, like gets really, good at, AI, right? Like what happens when Elon, who has never come into a market first or lost it because he comes from behind and just dominates this thing? What happens when, like these models get really, good and we go from those like two frontier labs to three frontier labs to like three frontier labs plus China to three frontier labs plus China plus all the open source models to like Meta, like all of that stuff coming online. you know, the GLM, like 5.3 stuff with like, you know, Baseten and Fireworks, like running all of this inference, right? Point. Like, yeah, I think, I think it's going to be a very interesting, 2027.

Nakul Mandan:

So everybody talks about compute as a second-order effect, but what else? Like, are there other aspects that are going to just— we're going to hit a wall and suddenly you just can't use this AI that we've unleashed?

Jake Cooper:

Well, I think it's like, it's like, permissions and like access and like keyring and, stuff like that. Because you can't just like unleash these things. Like a lot of people are just like, okay, like run it in a Sandbox with dangerously skip permissions. And like to some extent that sort of works, right? But like you wanna give these things wallets and you wanna give these things permissions and you wanna give 'em access to your, you know, like the whole like OpenAI thing. It was like everybody was like, here, please take my iMessage, take my emails and all this other stuff, right? And they're just like shoveling all this important information, right? Like how do you get to a point where you can partially dissolve that barrier to say like, okay, cool, you can have access to this section of emails or you can have access to $1,000 or anything else like that. And you can watch Stripe already do this with like Link or anything else like that. I think Coinbase is doing some stuff with like Agentic Crypto with, like X42, you know, stuff, right? And so like these things are, again, just like going to go and occur, right?

Nakul Mandan:

Are, the things that agents need from infrastructure that people historically haven't?

Jake Cooper:

I think I'm actually more on the side of like the, the things they need are not new, but it's just like the loops are gonna close a lot faster, right? And so internally we say that, software development has been solved, but software engineering has never been more important, right? And so things like, hey, listen, you know, I need you to build this feature, I need you to get it running in development, I need you to run it, get it running in staging, I need you to get it running in production under a feature flag, and I need you to go to roll that out. Up, right? And then go and watch Datadog or observability dashboards or anything to make sure you just didn't blow up, you know, this thing. And then listen to customer support tickets, right? You can automate that entire lifecycle. It's just completely automatable right now. And we've automated a lot of it internally. And then we'll go in and offer that to like folks, right? And so I think that like those things, you know, of the new things, there's probably fewer of the new things. they'll probably be very, impactful, but there's just so much low-hanging fruit right now that you can basically just say like, well, what was I doing manually and how can I never do that manually again in my entire life. And I think that's what a lot of the kind of folks over at Anthropic are talking about in terms of like loop engineering, like graphs and like all of these other things, right? Like it's, very important. It's, still diffusing.

Nakul Mandan:

So what does the Railway platform look like in 2 years from now? So what does the Railway platform look like in 2 years from now?

Jake Cooper:

Yeah, so I think the Railway platform, it looks like, almost like an orchestra in the sky building cathedrals. So I think incidentally a lot of the market is gonna move towards this, potentially new form factor. I think you're watching like OpenAI, they've got this Code named Dime thing. You've got, you know, Meta has been plowing money into like AR, VR for, like ever, right? Apple just like, you know, with John Ternus, I think is his name, the new, the new CEO, you know, they, killed the, Vision Pro or whatever. They're, moving their, stuff into like AR, And so I think that, you know, the, there's higher bandwidth mediums that are going to be more accessible. You have text, you have voice, and then you have like neural input basically, right? and so being able to manage an enormous fleet of like agents or anything else like that, I think is very, clear. And just being able to build whatever you want on top of Railway from wherever you are.

Nakul Mandan:

Just to make sure I understand that, are you saying that Railway will enter the agent, code authoring or no?

Jake Cooper:

Probably, right? Like we are already running like cloud agents for you in the sense where like you can just run like, Railway CA in the, terminal and it will go and pull in your OpenCode, your Claude, like whatever, all of those things and just run these things in the sky for you. Right. and, I think that a lot of development is just gonna go in and move that way because you're just not gonna have enough space on your, on your laptop. Like with the, with the way the hardware prices are right now, with the way that like everything else is, you're just blowing up the massive, amounts of complexity while simultaneously you're going through a compute crunch shortage on device. All of that shit is going to go and move to the cloud, right? You're going to get a re-cloudization in the same way that you had cloud compute in the early, you know, 2000s, late 2000s. Yeah. So, you'll see, you'll see a very similar kind of thing, you know.

Nakul Mandan:

So I want to shift towards your inner game. Jake as the operator, Jake as the founder. You've got all these management philosophies that aren't in any book. Like, were you— did you grow up first principle? Was that like— did you always have a disdain for conventional wisdom?

Jake Cooper:

Disdain is such a fun word. yeah, I, I wouldn't say it's like a disdain. It's like, it's one of those things where it was like I just always wanted to like go in and figure it out myself and then kind of go in and gut check my homework, right? So I was like taking apart electronics when I was like younger, right? Like I was like, I had an N64 that I like chopped up and like added a screen to it so I could like take it on like on the go or whatever. I think like it was always just like, how do you get the understanding for, oneself and then try and like see where other people were kind of going with it, right? So I've just always like operated, like that.

Nakul Mandan:

Is there a conventional wisdom that you've ignored that in retrospect you regretted? You're like, no, that conventional wisdom is— it was right. I, I was unnecessarily trying to reinvent that.

Jake Cooper:

I think like, you know, as t goes to infinity, I will be wrong about everything, right? Like again, because it's just, you know, disagreements of timing and, when you should do something or whatever, right? And like, you know, we've, prioritized one thing over another and then like, you know, within period of time we've been like, oh shit, like no, right? And, so I think it's, very easy to go back and say like, oh, listen, it would've been awesome if we had seen that like first version of like, you know, the GPT that like outputs code or whatever. you know, and we're like, that's gonna be the future, right? But like, time doesn't work like that, right? Like, and so you have to be, you know, of, your values, of your opinions, right? And then kind of be very, malleable to say like, yeah, I was wrong about that. Let's like move forward with this. Again, that's why I put myself internally with like 3 people, right? Because there's plenty of times where I'm just like, I think this is a terrible idea. And then people are like, no. And I'm like, all right, great, we're doing this. go, And then I will be wrong.

Nakul Mandan:

Where have you had to grow as a CEO over the last couple of years.

Jake Cooper:

Yeah, I think, I think like letting go is like a good one where you're just like slowly letting go of like more and more things, right? Like letting go of, you know, like sections of the business to allow people to kind of like own them more deeply. Cuz like if I'm just like, you know, holding over engineering, like, you know, it's like nobody can kind of grow and then make those mistakes. So like I've always just like tried to figure out how much I can kind of like let go and just various different, you know, things, right? And so just letting go is like going from engineer to operator to like business person, you know, all of those other things. And so just like letting go of all those, toys and, especially the ones that are fun that I really enjoy, like working on, like from an engineering perspective and all that other stuff. Like sometimes I just, I'm like, wow, damn, like that's, a really fun problem. And this problem that I'm working on today, it just sucks. It's just not a fun problem. you know, but you have to like go in and, do those things sometimes. And so it's just a lot on like letting go and just doing the one thing that will move the business forward for that week and just do that 52 times a year.

Nakul Mandan:

Are there aspects that you recognize today where you were bad as a CEO a year ago that you're good at now?

Jake Cooper:

Oh, I think, I think like the classic one for like founders is just like you don't realize how fucking bad of a manager you are until like you get better at being a manager. And I still think that that's a journey that I'm on and like learning and like, you know, now that we have our first kind of like group of like leaders internally forming, et cetera, you know, like learning to manage managers is different than learning to be a manager oneself, right? And so you just have to consistently put yourself in a spot where you were just like bad at something, right?

Nakul Mandan:

Can you give a tactical example, on this? Like, this is a fast— I think every founder CEO is going through this journey at some level. Like, what were you bad at managing? Was it communication? Was it holding people accountable? Like, give us a tactical example.

Jake Cooper:

Some— sometimes it's communication, sometimes it's feedback, sometimes you're like too direct, sometimes you're not direct enough, right? Like, you know, we use the concept of like right internally, right? Like, it's like right Trust, right? Precision, right? Clarity, right? Whatever, right? Because, you know, depending on the scenario, there's just a different amount that you can go in and apply, right? Like, you know, I've been working with a bunch of designers recently, right? And designers are like wonderful, people. They see, they see everything. And we have this concept, and, internally, called progress pics. And so it's just, you're working on something and you go and you share it, right? You're just like sharing it, right? And like engineers, like they will, they, they're, very precise. but they miss things once in a while, right? And designers are very different, right? And so having to go and learn those lessons to be like, oh no, like designers have like an enormous amount of perception. Like they just, they pull in so much input from all this thing, so they don't miss any of that, right? Okay, cool. We should do like design crit, which is like a standard thing where you're just like, hey, you bring all the designers together and you say like, hey, like let's talk about these things, right? Let's give really, clear feedback versus you share it and you get feedback from various different people, which works really, well for engineers because they can just be like, yeah, I'm not going to listen to 80% of that shit. I'm just going to go in and do this, right? so, Yeah, I think that those are just things that you go in and learn. And, even in that statement that I just uttered, there's probably designers like listening, be like, he, I don't even know if he knows what he is talking about or whatever. And it's like, they're probably, they're probably right. And so you just have to have that like learned mindset of you, will consistently be the first person on the dance floor. You will consistently be the person that like, you know, has to figure out some sort of new problem and you have to surround yourself with people who are gonna go in and be able to go in and do those things that you're just like bad or weak or whatever at. And like either don't want to or just don't have the time to like get exceedingly, good at, right? And I think like, you know, that's why you have to let go of the things you're, really good at because you will just like, convince yourself that like, that's totally fine. And like, it's, totally good to just like leave that thing there, right? But you will end up, you will blink and in retrospect you will realize you were bottlenecking the company, right? You'll bottleneck the company just by doing that, right? And, if you want people to be successful, if you want your investors to be successful, the people that are work— you're working with, if you want them to be the best, version of themselves, you need to figure out how to get out of those bottlenecks. Like as soon as possible, you know.

Nakul Mandan:

Is this something that you're working on right now on making yourself better as a leader, as a CEO?

Jake Cooper:

I think it's always like, you know, precision of just like, I think giving people the right feedback at the right time to go in and get them over, you know, their, hump or the thing that they're working on or anything else like that. I, I personally believe that it's like, it's difficult. Like you can only get better at it. You're not going to like get like perfect at it basically. I, I guess you're not going to get perfect at like anything. But like, I think people are very, difficult, right? And in the early days I was like, wow, this sucks. maybe I don't want to go in and manage people or whatever, right? And then I was like, unfortunately, like, started the company, have to do this. Like, there's no other option, right? so in for a penny, in for a pound. And then I learned that like, you know, people are so wonderful and they're like, there's, very clear reasons why people go in to do specific things. And if you can kind of like bridge the communication boundary and like talk with them like earnestly and, honestly and just like land to hand that piece of feedback or be really clear or anything else like that, you can kind of close that loop. You can turn it over with like pretty much anybody, right? and so trying to just like get quicker at that, right? Like precision is the better part of excellence. and then another part of it is speed, right? And so you trade speed for precision and you just try and get really, good at those two things.

Nakul Mandan:

What's a mistake you still think about?

Jake Cooper:

It's a mistake that I still think about. Oh man, these are just like baked into like lessons or like habits or anything else like that. I used to not plan any of my schedule cuz I wanted control over my own time. And then I realized it was just an interrupt factory of just stuff that was happening all the time. And then, you know, I, I've regimented myself and now I'm just like, oh, this is awesome, right? Like I have so much free time because I've just like put all of these things like right in. I mean, there's the classic one of just not, you know, being direct or like, you know, not letting go of somebody who you just like know is just like, you know, those things. I think that those are, hard and like I think about that every single time when somebody is underperforming and it like, it goes back to the value. It's like, oh man, like do I tell this person this thing today or do I tell them that tomorrow. Acceleration. Figure out how to tell them today and figure out how to do it in an honorable way where it's reliable and it's direct. Because like, again, done right, great people who want to be great will appreciate that.

Nakul Mandan:

I mean, the company is growing super fast. AI is changing everything very dynamically, every— for everyone in Silicon Valley, but also you, you're growing super fast. As you're growing yourself, like, do you get exhausted? Like, is the intensity something that sometimes exhausts you, or you're just— I would imagine both can be true, that you're having the time of your life, but is exhaustion, burnout a thing for you?

Jake Cooper:

I hold the belief that burnout is— it's like an input-output mismatch. And so if you put in a lot of energy and you're not getting the output you want, that's how you're going to burn out. You're just like, I'm putting on this thing and it's not moving forward, or like, I'm working with this person and it's not working, right? Like, being taught that lesson again and again and again and again, like, that's how you're going to go into burnout. And so for me, like, if I even feel tired, I'm like, well, what's the lesson here, right? Like, and then I go and I relook at, kind of like very core key things like, you know, diet, exercise, caffeine, kind of stuff like that. Like, are those all like in line or did I like not get enough sleep for the last few days, right? And then I like go in and fix those and then I figure out what the next like kind of bottleneck is and I just like do, you know, almost like the— you've read The Goal?

Nakul Mandan:

Yeah.

Jake Cooper:

It's a theory of constraints, right? Your, physical being is like a theory of constraints, right? All the way down. You can always like be better at a lot of those things, right? And I've, only found that I've, kind of run into those things when I'm doing the wrong thing at the wrong time. If, I'm like doing the right thing at the right time, it can be arduously hard and I'm still just like, let's go.

Nakul Mandan:

But there are phases in the company, actually, just to double-click on this, where you probably are pushing something and it's not getting done instantly. It's not about bad timing. It might just be that it needs the right solution. How do you assess as to this is an execution issue or a solution issue rather than the wrong problem to work at?

Jake Cooper:

Dyson has this, theory where he's basically like he's making one change every day. So he's done 5,127, I think, prototypes, basically. And so if the thing isn't moving, is it you haven't applied enough force to it? Is it like you, gotta try something like different or whatever, right? But you always try and like isolate it to like one variable, right? And if you can drill it down to one variable, you can usually figure out what's wrong with that thing and you can usually figure out how to go in and move it forward, right? But there are like, I think like big initiatives too, right? Like as we've kind of like built a lot of this, what we're codenaming like Railway Express internally, we've just like made a bunch of like, I don't know, just like blunders in terms of like, there's so many ways we could have just done this like faster. If we were better at project planning or parallelizing or all of these other things, or like collaboration on some of this stuff, or like we had a conversation about like swim lanes today, right? you know, stuff that's like, you could have just done this. and so it's always just good to like almost like, and this is another reason for like to avoid, you know, 996, which I'm not telling people how to run their life or their business or, whatever, right? But I think that period of time, that 36 hours for the week where I can let my brain decompress and figure out where my blunders are for that week and then like come back the next week with like a clear head, a clear mind, and just say like, not doing that again this week. Fuck that up. We're not doing that, right? Like, okay, cool. Let's like move forward on that, you know?

Nakul Mandan:

So next up is our quickfire round. You ready?

Jake Cooper:

Yeah, sure.

Nakul Mandan:

Do it. Worst hire you ever made?

Jake Cooper:

Worst hire you ever made? background person. you know, one who has, had stuff in their background. yep. So that would, that would be bad.

Nakul Mandan:

Red flag you never ignore when recruiting?

Jake Cooper:

Like lack of drive, basically. Like, you know, asking about like things that are, you know, typically associated with like passengers, tourists, all of those other things, right? you know, those, are the ones basically. Not wanting to do hard work. It's just never ignore that one.

Nakul Mandan:

Most overrated startup advice?

Jake Cooper:

I think most, startup advice is overrated. would be my, would be my take actually, because like you have to plot your own journey. You have to figure out all this shit yourself and what works for that person. You know, I have wonderful mentors, right? And not that they're giving me bad advice or, whatever, but like the, dynamics just change. Things are different or whatever, right? And so you need to basically take an advice, pull in a lot of those inputs. It's like reading a business book or whatever, right? Like I, I tend to read a book and then I try to forget about it and then the advice is like applicable because it was like imbued into my limbic system, you know? and so that's, I think most advice is overrated until you hear it multiple, times.

Nakul Mandan:

Something you believe about building companies that most people will disagree agree with you.

Jake Cooper:

Almost every person can be excellent if they want to, and if you can bridge to them and figure out how to be, be excellent.

Nakul Mandan:

Data centers in space, yes or no?

Jake Cooper:

Yes. I don't think the math works right now. I haven't personally worked it out myself, but I think that there will be a way to go in and figure it out. It makes sense.

Nakul Mandan:

Best thing about being 31 and running this?

Jake Cooper:

Being neuroplastic enough to, get better every time, but also starting to develop a little bit of wisdom, right? And so you can start giving advice, you know, as you're kind of in this period where you can take inputs and you can give inputs and you can feel like things are turning over. Like, and I can give, I can give advice, and again, most startup advice is overrated, but I can start like helping people earlier on in the journey. That's like so amazing. But I still like get like really, great advice from like the greats or mentors that I have or anything else like that. Right. So it's like the time—

Nakul Mandan:

sweet spot.

Jake Cooper:

Yeah.

Nakul Mandan:

Worst thing about being 31 and running?

Jake Cooper:

You're so fucking dumb. you know, like you're just not like you're in, and if you do your job correctly, you will be dumber every year. You're just like looking back and you're I was like, holy fuck, what, do you mean? That's what I thought at that point in time. That's, awful. Like what? You know, we, hope for so much better, but I think that, that's, that amplifies growth, right? And you're just growing all the time, but like you go back and you're like, really, man? Like you didn't learn that lesson fast enough to like go and just avoid all those things.

Nakul Mandan:

I know you're a first principles thinker, but which CEO would you shadow for a day?

Jake Cooper:

I think it would have to be like Sam over at OpenAI. I just have to imagine that, is just like, there's just stuff happening every single It's like everything, you know.

Nakul Mandan:

Last question, a young founder watching this wants the life you've built. What's the one thing you tell them that this journey actually requires?

Jake Cooper:

Yeah, so, I would first say that you probably don't— maybe it looks like you do, but like it's a lot of work. you know, you have to really, want like this specific thing. I would say that like really, decide like is this company, is this thing like the purpose that you want to embark on, right? Because I think that if you do this correctly, you will be around for 10 years. 10 years, right? And you only have like 2 or 3 10-year blocks in your like adult lifespan before you kind of just like peak and then fall off, right? Like it, just gets more difficult as you get older, right? And so like be very, deliberate about your time and only really kind of like do things if you think you, can do them for like a long period of time because like pretty much everything in life, that's great, like compounds, right? And if you can get that compounding period over time, then just like you just have an enormous, impact versus these kinda like smaller things, you know, right? Like I think a lot of people, like, they, rotate way too quickly between like companies and stuff like that. And they're like, I'm gonna work at this company, I'm gonna work at this company, I'm gonna work at this company, I'm gonna go to YC, I'm gonna like start the startup and then whatever, right? It's like, okay, but like, what do you want to, what do you wanna do? Like, what's your purpose, right? You didn't, you didn't state any like specific thing there other than like almost like ladder climbing, right? So try and figure out like where you believe that you might be meant to be at this point in time and then take a shot, like figure out, make a mistake, like decide, yeah, I'm going to go in and try and do that, right? And like learn from that mistake. And then I think you'll turn it over a lot more quickly than, trying to follow the, conventional path. But again, I'm, very biased because I'm like, you know, you said first principles. I'm like, don't love that word or whatever. It's like, I think it's overused, but is correct, in terms of like, yeah, it was just like, I want to learn it myself. A lesson, like, earned, a lesson learned, a lesson burned. Like, I'd rather be burned, you know?

Nakul Mandan:

This was great. Thank you so much.

Jake Cooper:

Cheers, man. This is awesome.