Pinned Tweet
My life's goal is to make this obsolete -
xkcd.com/303/
When we started on the Argonaut journey, we got 1-2 friendly customers while building out the MVP. But then, we stopped selling.
I was delaying launching with reasons like “we don’t have self-serve yet” or “this feature will make for a nice launch”. It took a while to come to terms with it, but it was fear of rejection for someone trying to sell something for the first time.
We are doing things very differently this time around, and learning from previous mistakes.
✅ Picked a problem that is narrow, and the value proposition can be conveyed in one sentence.
✅ Put up a website and a waitlist before development to gauge interest.
✅ Continuously selling - everywhere, all the time.
Everybody wants builds to be faster. Everyone wants costs to go down. The value couldn’t be clearer and we are seeing fantastic results.
Earlier today, a customer got setup within ~2 minutes and this is what they had to say.
I think we are onto something here 🔥
WarpBuild retweeted
Chris from @typecraft_dev switched his github actions to WarpBuild with a one line change. the build was done before he could even click into it
Chris from @typecraft_dev switched his github actions to WarpBuild with a one line change. the build was done before he could even click into it
WarpBuild runners can run in your own AWS, GCP or Azure account. we manage them, they run in your VPC. static IPs, any region
WarpBuild retweeted
My coding agent took 9 attempts to fix one failing test this week, and the model was never the slow part.
Every attempt ran the same loop: push, wait for GitHub Actions, read the log, push again. The model replied in seconds. Each CI run took minutes.
Once an agent is blocked on a green check, CI time becomes its iteration time. More AI code means more commits, and more commits mean more runs sitting in the queue.
WarpBuild's public benchmark (spotify-player, Rust build, 4 vCPU, median of 3 runs):
GitHub-hosted: 4m 34s
WarpBuild: 2m 43s
Across ten agent attempts, that's 27m 10s of waiting instead of 45m 40s.
The switch is one line:
- runs-on: ubuntu-latest
+ runs-on: warp-ubuntu-latest-x64-4x
Workflows, actions and YAML stay as they are. There's no job concurrency cap, so parallel agents don't queue behind each other, and macOS runs on M5 Pro with unlimited concurrency.
WarpBuild retweeted
my github actions are pretty slow and most of the times stuck in queue, so
i started finding alternatives. found a few of them with warpbuild giving best performance
i ran same rust build, same 2 vCPU size, CPU pinned at ~100%
speed: around 2x
cost: less than half per run, so pretty happy
🤝 Paid partnership
Replying to @kentcdodds
@kentcdodds's agents kept hitting CI concurrency limits. so he switched to WarpBuild
Use WarpBuild to run CI in your cloud without any overhead.
AWS, GCP, Azure are supported.
If you're running GitHub Actions runners on Kubernetes with ARC, you know how much maintenance comes with it. @WarpBuilds BYOC is an interesting alternative. No control plane to manage, support for Windows and custom VM images, and potentially lower costs without worrying about bin-packing on K8s nodes.
They also offer fast hosted runners, including macOS.
Worth checking out if you're spending too much time (and money!) maintaining CI infrastructure instead of actually using it.
Get $50 in free credits: warpbuild.com/start?utm_sour…
audit logs are live on WarpBuild
every change to your org: who, what, when, and whether it went through. denied requests and staff actions included
write-once, so nobody can edit or delete them, us included. kept 13 months in your data region
filter, export to CSV, or pull from the API
WarpBuild retweeted
why didn't i find this tool earlier??
i added warpbuild runners in my github actions workflow and my build went from 91s to 52s.
it costs less than half as much per run.
all i did was replace the runs-on value with:
runs-on: warp-ubuntu-latest-x64-2x
same build, same machine size. warpbuild is faster and cheaper per minute ($0.004 vs github's $0.006), so i'm paying less and waiting less
coding agents have increased my development speed, which means more code changes and more PRs to check.
pretty happy with this change.
warpbuild.com/start
🤝 Paid partnership
we've been busy. here's what shipped at WarpBuild
1. audit logs
2. action debugger v2
3. metrics export interval
4. tighter GCP cloud connection
5. security patches
WarpBuild runner metrics can now go straight to your observability tool
CPU, memory, disk i/o, network + logs from every CI job, over OTLP. works with Datadog, Grafana, Honeycomb, New Relic, SigNoz and any OTLP backend
Settings → Telemetry export
WarpBuild retweeted
It's a good thing @WarpBuilds doesn't run on firecracker.
Dodged this one.
WarpBuild retweeted
Thank you @WarpBuilds for sponsoring this episode. Faster, cheaper CI this way 👉 kcd.link/warpbuild
macOS sandboxes on WarpBuild now come with a full desktop
one command opens the sandbox's screen. build your iOS app, then run it on the iPhone Duo simulator, all on a cloud Mac
wb sbx vnc <name> --open
building for iPhone Duo? build and test it on WarpBuild
Xcode 27.1 + the iOS 27.1 simulator are on our macOS 26 runners. point xcodebuild at iPhone Duo
MacOS sandboxes on WarpBuild
You can pause, resume and pick up right where you left off
No compute charges while it's paused. Billed per second when it's running
full demo 👇
macOS sandboxes are now on WarpBuild.
Spin up a macOS VM with one command, with Xcode and the usual iOS tools already installed. Copy your code over, run your build and tests, then pause it when you’re done.
While it’s paused you don’t pay for compute, and your home directory and tools are still there when you resume.
Billed per second.













