Railway's Developer Infrastructure Focus - Coconote

There's only two modes. It's hard problems doing hard things. We want to make it as cheap as possible in both time and money. Saving you a ton of time to go and spin up these things and then only paying for what you actually use when you're going and doing that. We want you to spend as little time as possible on the kind of configuration, drudge work, etc. … You can just do it yourself. ... And our conversation gets into how they solve the hardest problems in dev tools and the importance of back-end infrastructure in AI native software. In the age of AI, there's a very very material risk of being out executed by your competitors because they are moving so quickly. … I think this will be fun. really quick for people who don't know I've heard railway described as solving the hardest problem in dev tools what what is that actually and then what is what is railway yeah so I think it's like the intersection of a lot of hard problems and so that's what makes it quite difficult so we are really … So there's a lot of hard problems in between whether they're user experience problems, infrastructure problems, configuration management problems, versioning, deployment, rollouts, kind of stuff like that, live infrastructure. So pre- railway, what was I doing to do all this stuff? Like what did the world kind of look like? ... So we were kind of the thing that I started with was like databases right? And so databases are actually pretty hard like stable storage ends up being actually quite difficult on the you know almost like iceberg of like complicated infrastructure things right because you know people have data in their databases and if you lose the data in your database then you have a terrible time or whatever right so I started with that um because I … I just want to move a lot quicker right and so our as our capability kind of gets significantly better and better and better and those people wanting to deploy faster and faster and faster we're reaching this kind of inflection point where most people are actually almost capitulating out of that like older mechanism of like building software and moving into the realm of the thing that we've built right so That's super cool because I think for a long while we're building a thing and you know we we go and like pitch it to like VCs and they'd be like well like you know what problem is it solving? It's like well software infrastructure sucks right and you're like yeah but like you know tell me more about this thing right and now it's people want to move significantly quicker and in the age of AI there's almost like a very very material risk of like being out executed by your competitors because they are moving so quickly and if you don't have … sp internally and so they ran into essentially this like platform team edge of like this team potentially internally that you know manages all of the this tooling they've exposed them in various different ways that ends up being cumbersome and that's through no fault of the platform tooling that's like actually the under the underlying infrastructure like the the infrastructure that they are trying to abstract is very very difficult and in my opinion is like almost unobstructable. … So it was like really really poorly fleshed out in in general.