Railway raises $100M to build data centers and capitalize on 'agent speed' software development

our Lambda Lightning round and bring in Jake Cooper from Railway. He's the co-founder and CEO, also a TV sponsor.

There he is.

Jake, good to see you. How you doing?

Get that going.

How are you all doing?

We're doing fantastically.

So stoked finally have you on the show.

Yeah, great to have you here.

Give us the update. What happened?

Cool. Uh yeah, I mean we raised uh a bunch of money. So uh raised $100 million round to build data centers. um and make software just way easier to use because the definition of a developer is melting before our very eyes.

It's melting, you know. Um so anybody can ship code which is crazy. Um [laughter] we're making it trivially easy for anybody to go and do that. So it's it's a really exciting time to be uh you know in the business and also just you know on the planet.

Yeah. What uh how is that actually playing out? We've talked to some folks about how there are designers that there's one idea that AI will create will take a 1x engineer and turn them into a 10x engineer and that's probably true. Uh but there's also this idea

it's going to be like a 10,000x 1000 like we have people internally moving so so quickly. I've like rebuilt a bunch of the internal organization so we can move at what I'm calling it's super cringe but I'm calling it agent speed [laughter] because people are just moving way way faster now. Um, so like you get a 10,000x engineer. Yeah. Sorry.

But then but then there's also the zero to onex engineer where you take a designer and then they're able to actually ship production code or a manager or a sales development representative. Anyone can build a tool. Uh which one is more of more central to the current railway pitch when you're talking to customers?

I I think there's like resonance across the entire space. So we have like larger companies who are you know they're spending six figures. they're they're building something like of deep consequence like critical critical critical for their business, right? And then you have like teams of, you know, a thousand essentially all kind of like vibe coding and they're building like really really kind of smaller things. Um but they're really really important either for client infrastructure or internal tools or anything else like that. Um and you know we've got everybody internally now on like a cloud cloud code subscription. Um you know in including our recruiting, our operations etc. Um there's just the boundary of your ambitions needs to be deeply deeply kind of re-evaluated just with how good these tools are getting. Um and we're excited. What's your what's your personal framework internally and kind of like guidance or how how are you and the team thinking about like when it makes like uh cuz right now I feel like there's this obviously so much excitement you can build you know an internal tool in an hour or something like that but every time you build a tool you're creating like some type of like ongoing uh maintenance burden something like that

tech tech so but yeah so what what's your framework of like hey this should be something that uh that we just should build versus you know finding a point solution or or even using something as like a Google sheet or uh at the at the uh kind of low end.

Yeah. So, so, um, we use notion a lot internally and I think lots of like deeply flexible tools are going to be really really key because now that the agent can kind of go and almost like orchestrate these things in general, um, you can kind of compartmentalize very very flexible tools together, right? And so I think that when you're thinking about kind of the build versus buy boundary, right? Um, it's a lot on kind of pushing the context into the tool so that you can access a lot of the context and take advantage of it, right? Um so I think the the burden for uh almost buy goes up a lot um because the ambitions and the capability of you being able to build something that's purpose-built internally is actually so trivial now right um like historically I've I've built a lot of internal tooling to allow us to do project management releases etc. Um and now I can actually just kind of push the boundary on a lot of those things. One of my hot takes is that like um you know ERP software right? of course with it, right?

We love ERP software. So ERP software, right? Only useful at scale uh to build an organization to manage resources or whatever, right? Um now you can kind of build those like organizational structures way way earlier, right? So the way that railway runs um has historically been different than organizations. And so I think what you're going to see is almost this cam cam explosion of various different ways that you can run your organization, which I think is super super cool. So um unfortunately like I'm an engineer so I have engineering brain damage. I think everybody should be building kind of their own stuff. Uh you know um so yeah.

How has your thinking on product management evolved over the last year?

Oh man. Um yeah I I had this this tweet that I I was like oh yeah engineers are going to turn into PMs like faster than you know PMs can turn in into engineers. And I think like a week ago I was like that's completely wrong. Um [laughter] you know like

I like I like that you were willing to to just be like yep bad take.

Oh yeah. I mean, you got to you got to take your lumps, too, right? Like, um, I I don't mind like your take can be awful, but if you're going to stand behind it, just like almost like go down with the ship and be like, "All right, the [laughter] take was awful." Um, you know, we're going down. Um, you know, don't delete the tweet. Um, so yeah. Um, but yeah, I think I think the there's only three things essentially left if you assume that the agents are going to kind of be able to do pretty much anything, right? Um, it's taste, agency, and structure, right? Um, so when you think about almost anything that you can go and build, those PMs have a lot of that context, right? And so you don't need to be a Rust engineer to write Rust anymore. You don't need to be a JavaScript engineer. You don't even have to be an engineer, right? Like we're almost doing this like like double shift, right? Uh where you have like knowledge workers, programmers, engineers, architects, and scientists, right? Um based on how nebulous your work is, right? And I think that if you're a knowledge worker, you should probably actually just be like, yeah, I can just engineer things now, right? What it takes is kind of a bit of critical thinking, a bit of kind of like, okay, how will this system kind of roughly fail? Um, and now you can do it in kind of a lower stakes environment. we hope to provide um with railway is actually those kind of uh like bleeding edge industry best practices like really annoying stuff like VPC peering uh like you know versioned uh infrastructure uh so that you can move quickly but also safely and build these kind of bigger and bigger structures over time. So I think I think the role of the PM evolves a lot and and you're going to see a lot of these kind of almost like merging or melting of these kind of roles together, right? Like software engineers will have to become PMs because now it is completely trivial to just go and and write code, right? Like I wrote like 5,000 lines of of code yesterday, right? Um and I spent most of [laughter] my Yeah. I mean, and this is like between meetings, right? Um and and I spent most of the time not actually um like auditing the code itself but auditing a lot of the logic going in working talking about the architecture all of those other things right um so I think again the the burdens or the the boundary of your ambitions needs to be deeply deeply um pushed and so if we at railway can kind of almost unbburden the developer so that they can think about those bigger and bigger things our hope is that you can build those kind of almost like massive massive structures right like if you look at the Golden Gate Bridge in this current day and age. My first uh kind of response is like how the do we build that thing? It's so big and so impressive, right? And so now you actually have these wonderful tools that push the boundary of that. So you can build these massive things, right? And it's super exciting to just, you know, you know, I was telling this to an investor, which is probably the wrong thing to tell to an investor, but I was like, even if we get schmucked here, like what a time to be alive. And they were like, [laughter] cool, dude. All right. So, um, we're gonna pass. Um, but yeah,

that's wild. [laughter] Uh uh how are you thinking about

No, I do I do think that like uh I mean obviously you know play to win and and uh you know but it's very it's very refreshing versus the I know exactly how the next 5 to 10 years will unfold day by day. It's like no, no one knows like the space is moving so quickly. You need to really deeply understand like who's signing up for your product, who loves your product, why do they love your product, how do you,

you know, expand your product and continue to be, you know, critical infrastructure uh for those customers. So,

we have some questions from the chat. Uh

yeah, yeah, sorry. Uh what were you about to say?

I was just going to say, yeah, like I I think that like during this period, you should kind of take those risks, right? Like it's it's almost a deeply uh almost like risk by default period, right? And and I think that you know um there's like what does Elon know that that probably other people don't know right and I think you've heard like Peter Teal talk about this too and it's almost the calculus of risk is not it's not just purely there right uh risk is less risky than you you actually think right so I think people should kind of push those boundaries and if we can kind of create those safety nets for people in general um I think we can we can build wonderful wonderful things and as a side note I think we've made uh risk and failure a bit too expensive in our current society and if we can re like rewind that essentially we can open people up to like sweet success and a bit of bitter failure because when you take risks some sometimes things don't work out but that forms a baseline experience that allows you to kind of drive these unique kind of tastes forward. So to the chat to the chat uh we have to ask you about your bike

your and says uh your motorcycle liability clause on your term.

Yeah we have key man insurance. Um yeah I'm I'm kind of like you know I'm counting the days until somebody really asked me to like go up and give up the bike. And and it's funny because people people ask me like oh my god you ride a motorcycle. Like I I feel like I should get into that. I was like my answer is like absolutely not. I do not do that. Like I I've been riding dirt bikes since I was like 13 basically. And so and so like I've had I've had a bike for a while and I've kind of worked my way up, right? Um you know uh yeah don't just don't be stupid.

No say the riskiest thing you can do is not take any risk. Get on the

motorcycle ride.

My Sunday ride is uh basically I I live near the enrichment so I basically just like go up and drive through the Prescidio maybe in the marina. There's that like nice uh coffee uh like truck basically. Um I I go to the gym in the marina. Um they've got a sauna. It's wonderful. So I just kind of

Do you ever try to find like a super steep, you know, classic SF hill and go a little too fast [laughter] on the way up?

No, those those days are like behind me. Uh because uh yeah, I I I promised the investors that like I wouldn't do anything stupid. So the unfortunate thing about the bike is like it's a it's a nice bike in it's like the bike I wanted growing up,

but I don't ride it like like it should be.

What bike? What bike?

It's a it's a Ducatti like the kind.

There we go.

There we go.

Uh we have another question from the chat. Everyone's loving the bookshelf behind you. Uh, give us some book recommendations. What's on the shelf?

What's the most underrated book?

What's the most underrated? What's the favorite book? What's the one with the most uh highlights and notes?

Well, this is actually my wife's side of the the bookshelf. Um, so I'll go off things that are maybe not on there. I know. I do see I think the Steve Jobs book is somewhere up there, which it's pretty good. Um, one of my favorite books actually is uh getting to Yes. I don't know if you yall have read it. Um,

so [clears throat] is negotiation. Uh it's a book essentially on compromise, right? Um and so essentially it allows you to uh break down problems in a way that uh is not kind of me versus you, but it's me versus the problem, right? And so one of my favorite analogies in there is like you know we're splitting an orange, right? Um we could split it down the middle essentially and uh you know I could like you could have half and I could have half in general or we could figure out what we care about in general, right? And maybe I'm making mixed drinks and maybe you're taking your kid to the soccer game, right? Or anything else like that. And so in in essence I want the peel and you want the whole orange, right? we can both come out on top roughly right and I think that figuring out ways that you can almost like have a solid compromise with people is is deeply deeply important um you know on that one so I like that book

love it

uh back to back to AI and how it's changing software development uh database selection uh database migration switching something as uh ingrained in an organization as a database has historically been very difficult are AI agents making it easier are you actually seeing companies move move around from one database to another, maybe based on cost, maybe based on features, is is the is the fact that a particular LLM is in love with a Reddus or a MongoDB, like is that going to be relevant in the future?

So, um I'm unfortunately a little bit like too Daario pill in terms of uh like the the fundamental laws of physics, right? And so I think it's it's actually probably the hardest thing to change about your your system is uh the database roughly, right? Because you have that storage, you have all of those other things, right? Like this is fundamentally also why we build data centers, right? Um one, you need more production roughly in the world and we can start with data centers and that's that's wonderful. Um but two, it's actually a really solid mode in terms of um you know, you can't just will into existence a data center, right? Like and at some point um potentially down the line like those things will get more and more expensive just because of the cost of labor, but like the cost of capital is decreasing, right? Um and so I think the similar thing kind of applies roughly with data centers if you kind of move roughly up the stack. And so we actually started building uh or sorry databases. Uh we actually started building databases like we were originally just a you know click a button get a Postgress instance or Reddus instance or anything else like that and we've slowly worked our way back up right up to building you know web apps. Um you know we'll launch a CDN at some point in the future. Um stuff like that. U so I I think that when you look at that lens of like fundamental law physics I think the database is like a very very important part of your your kind of rough calculus. Um, and I think the agents are making it easier to go and uh, almost manage those migrations. But, you know, when I was talking about those 5,000 lines of code, um, you have to think about kind of what you are roughly building, right? And so, you're still going to need to audit it. And that's sort of why I say, you know, moving from engineers to like architects, right? Um, where you're thinking about that kind of bigger bigger vision. And so, you know, um, as as Jordy said earlier, like I don't think anybody can roughly pick uh predict the the future for a lot of these things, but I do still think that fundamentally there are going to be um requirements for how do you integrate the things that exist in the physical realm. So, roughly around those data databases to storage to data centers um and with with kind of your expected reality basically, right? Um, and the hope over time is is um that railway can actually help you do that, right? So we do this on a software level. Um, and we've built tooling internally to help build data centers. And so we've actually virtualized a bunch of those racks. So now you can almost drag and drop those servers into them. And at some point we can get to a point where we're working directly with the super micro API. You can drag and say, "Hey, I want this data center of this size in general." You'll create POS for that that and say, "Okay, cool. Hit a button. It'll cost you 25 mil or something like that." [laughter] And and that's one click check out for data centers.

One click checkout for data centers and stuff like that. um and and moving towards a world where you can almost cuz it's easier to build things on the internet, right? Like it's just it's a fundamental like fact, right? Um if you can simulate it and simulate that small section of it and you can create almost APIs to reconcile like this is what Terraform did really really well like expected state versus desired state. Um and if we can do that with your intention as you kind of move through time and give you those safe kind of primitives to say okay cool like I want to make this change. Let me fork my entire environment. uh you know get copy on right storage for for everything make some small change I make it in production it merges and you deploy it instantly then you're kind of moving at agent speed right um and that's that's what we think a lot about in terms of building the next fundamental layer of of software development and then physical production beyond that

you were at Bloomberg did you get to work directly on the terminal

uh yeah a little bit um I bounced around a bit at Bloomberg I was only there for about six months um but it was it was a really really cool cool time to kind of work on stuff like that. Um I was working on a bit of the internal tooling. Um so I got a couple commits that I think are probably still inside of the powering the bots podcast making Joe Weisenthal's job easier. We love to see it.

Putting Joe Weisenthal in agent mode.

Yes. [laughter] Hopefully. Well, thank you so much for taking the time.

Congratulations to the whole team. The chat loves you. Come back on again soon. And thanks for thanks for having us as sponsors. like we're just we're big proponents of like rich taste and like we believe that TBNN has a rich taste that y'all are sharing with the world and we're just happy to be here. So have me on anytime. Love to to join and hang out.

Thank you for making this possible.

Awesome.

We'll see you on for the C very soon I'm sure.

Bye.

Cheers. And in case you've been living under a data center, Railway Railway simplifies software deployment. Web apps, servers, and databases run in one place with scaling, monitoring, and security built in. We're very happy to be partnered with railway. Um we have Joe Weisenthal coming in into the

He said he was running five minutes behind so we can get back to the timeline.

Quickly let me tell you about 11 Labs. Build intelligent real-time conversational agents reimagine human