, and we couldn't be more excited about what's ahead. We'll get into what that means below. 

Of course, just because we’re doing a big announcement, doesn’t mean that the shipping stops. This week we’re bringing you an architecture view for the projects page, and focused PR Environments that only deploy what changed.

Railway closed $100 million in Series B funding

Over 2 million developers have shipped on Railway. Tens of thousands of companies, from weekend projects to enterprises, run production with us. Thank you for being part of this journey.

So what are we building? We want infrastructure to disappear. You should only have to think about making something great, or closing your laptop and going outside.

We call it an Intelligent Cloud, and if you've tried the 

 we shipped earlier this month, you've already felt pieces of it: your coding agent deploying to Railway without you leaving your editor, pulling logs/metrics from your running services and working with environments.

We want Railway to feel calm, and if we do our job right, you get more time. Time to build, time to live, time to do things worth doing. And when you come back to the keyboard, you bring all of that with you.

 if you want the full vision. We also made 

 about what shipping peacefully feels like.

See your project's deployed services at a glance

The projects page now has an architecture view. Think of it as a minimap for your infrastructure, giving you a quick mental model of what's deployed where.

Only what changed in your pull request gets deployed

If you're working with a monorepo or multi-service project, you've probably felt the pain of opening a small PR and watching Railway deploy everything. Changed one file in your frontend? Here comes the backend, the workers and the databases.

Focused PR Environments bring some intelligence to this. Railway now checks which files changed in your PR and matches them against each service's watch paths and root directory. Only the services that actually care about those changes get deployed. If service A changed and service B references it (say, via 

), service B comes along for the ride. Everything else sits tight.

You get full visibility into what's happening: the canvas shows which services were skipped, and the GitHub PR comment gives you the full breakdown. If you need to spin up a skipped service anyway, you can do it manually with one click.

This feature is in beta, so we’d love to know your feedback. Give it a spin and let us know what you think on 

We shipped autocomplete for the raw variable editor, making it easier to reference other services and variables

We shipped service config editing from the CLI. You can now set source branches, PR environment settings, and more without touching the dashboard. Great for 

automating PR environments with GitHub Actions

-service-config $SERVICE_ID "source.branch" "$BRANCH_NAME"

We shipped audit log events for workspace member changes. Previously only project member changes were tracked. Now 

We shipped numeric comparison operators for log filtering. You can now filter HTTP and deploy logs by things like "latency > 500ms" to find slow requests, or "status >= 400" to surface errors

We fixed the "View in Context" button in logs. Previously, clicking it would cause the log display to flash and fail to scroll to the right place due to a conflict with autoscroll. That's now resolved

We shipped flattened display for nested JSON in logs. Nested objects and arrays are now expanded and clickable for filtering, instead of showing as raw JSON

We improved error messaging for invalid domains. Instead of getting a generic "problem processing request" error, we now tell you when the domain is unsupported

https://res.cloudinary.com/railway/image/upload/v1769089616/blog_-_featured_image_lj2ft7.png

$100M Series B,  Architecture View, Focused PR Environments

Changelog #0274