@_lopopolo @charliermarsh If GitHub Actions, yes there are path filters. We shared some skills with 70+ patterns for CI optimization like that.
starsling.dev/ci-speedup
English
Daniel Worku
95 posts
Daniel Worku
@dbworku
Co-founder @starslingdev 💫 Previously NFLX, META and many startups
Katılım Kasım 2012
608 Takip Edilen744 Takipçiler
in the spirit of @charliermarsh's uv sizeminning PR, some free money for those who want it.
> is there a way we can optimize ci to be less expensive in terms of minutes? for example, the tf ci should really only run on prs if like tf code or package cvonfig changes, right?
English
Daniel Worku retweetledi
Kubernetes is the wrong primitive for diverse workloads x.com/i/article/2082…
English
@20thr @daytonaio @blaxelAI @modal @novita_labs @e2b @ivanburazin @bernhardsson @junyu_huang @mlejva @namespacelabs Totally! Can we get a concurrency bump?
English
@dbworku @daytonaio @blaxelAI @modal @novita_labs @e2b @ivanburazin @bernhardsson @junyu_huang @mlejva Would love to get your results with @namespacelabs
See computesdk.com/benchmarks/san…
English
Once we decided to migrate off of AWS, I set out to answer a simple question: which sandbox actually finishes real tasks fastest?
The results for our workloads:
🥇 @daytonaio
🥈 @blaxelAI
🥉 @modal
4) @novita_labs
5) @e2b
StarSling (YC X25)@starslingdev
English
@samraaj @daytonaio @blaxelAI @modal @novita_labs @e2b @ivanburazin @bernhardsson @junyu_huang @mlejva I haven’t explored memory snapshots. We have strict isolation requirements and only use ephemeral machines.
English
@dbworku @daytonaio @blaxelAI @modal @novita_labs @e2b @ivanburazin @bernhardsson @junyu_huang @mlejva i recently did something similar. Did you try out modal in memory snapshots? Bc that made my overall completion/spin up times way faster
English
Finally realest benchmark in sandboxes space!
In general, love the trend of benchmarks going beyond basic and overly general workloads and focusing on real world applications.
StarSling (YC X25)@starslingdev
English
@thdxr We just opened our benchmarking repo for real-world developer and CI/CD workloads.
@thdxr what work do you want do in the sandboxes? Can add a real-world test suite similar to these
#realworld" target="_blank" rel="nofollow noopener">github.com/starslingdev/h…
English
one month to go
gotta get the agents in tip top shape so they can keep workin without me
English
@t_blom @NotTomBrown Congrats Tom! They are extremely lucky to have your wisdom, judgment and candor.
English
Personal update: I'm taking a leave of absence from YC to join Anthropic. I'll be working with @NotTomBrown on the compute team.
Powerful AI has the potential to improve the life of every human on earth and, as we enter the early stages of recursive self-improvement, availability of compute becomes one of the most important issues to solve.
I'm excited to get started 🚀
English
@better_auth had 150K/week downloads when @minilek invested. One year later, it's at 4.7M/week and a part of @vercel.
Congrats @bekacru and team 🎉 One down, one to go.
Jelani Nelson@minilek
Excited to invest in two @ycombinator companies from this cohort, @better_auth the fastest growing auth solution @starslingdev AI agents for DevOps (“Cursor for DevOps”) @bekacru @dbworku @yonasbe
English
@crohr @yonasbe @starslingdev Wunderbar. Thanks for the honest comparisons @crohr! Good results for our initial launch. Will continue to ship faster and improve.
English
English
Daniel Worku retweetledi
StarSling Runners are now in GA! We did a ton of infra work to make our runners self-serve and are super excited that anyone can try them out now.
One customer who works with the frontier AI labs just signed up and migrated two of their most problematic workflows over to StarSling and saw an immediate 50% speedup in their jobs 🏎️
Every team should have faster, smarter, and cheaper CI.
StarSling (YC X25)@starslingdev
🗣️ Starting today, StarSling Runners are generally available! StarSling Runners are an AI-native, drop-in replacement for ubuntu-latest that make your GitHub Actions up to 6x faster and 13x cheaper 🚀 Everyone has the option of getting fast CI runners now. What nobody has is CI that adapts and improves itself. 🎉 StarSling has now run over one million CI jobs for companies like @better_auth, @mastra, Partcl, and more. If your team runs CI on GitHub Actions using ubuntu-latest, you're now just a one line change away from faster runs. Check out our Launch YC! ycombinator.com/launches/R7B-s…💫
English
English
In <2yrs @bekacru turned better auth into the #1 auth framework for typescript and one of the most popular and influential oss projects in the world.
He’s a world-class developer with impeccable taste. It’s hard to imagine a better home for him and his work than @vercel, or a better mentor than @rauchg.
Congrats to bereket and the @better_auth team!
Better Auth@better_auth
Better Auth is joining Vercel. Vercel shares our commitment to keeping auth open source, framework and platform agnostic. Joining forces gives us more resources and room to focus on building the framework that much of the community relies on. Together, we’ll also be able to invest more deeply in where auth is heading: a future where agents act on users’ behalf and need secure, scoped, and revocable access.
English
English
Welcome @bekacru to Vercel. It's a privilege to join forces with Bereket and Better Auth to further our Open SDK vision.
Developers deserve better auth, to serve humans and agents, built in the open and with taste. That's what Bereket has accomplished. Congrats! 🇪🇹🇺🇸
Vercel@vercel
The @better_auth team is joining Vercel to accelerate open source auth for apps and agents. vercel.com/blog/vercel-ac…
English
@crohr @yonasbe @starslingdev Also, @crohr can we get included in Disk I/O benchmarking? We should be near the top. :)
runs-on.com/benchmarks/git…
English
@crohr @yonasbe @starslingdev Hey @crohr! CTO here. x64 single threaded passmark looks right.
Will investigate the queue those results are actual slow. p50 is 8s as we use microVM sandboxes.
arm64 and custom images are only per customer for now.
English
Introducing Pulpie, models for cleaning the web.
70% of a typical HTML page is ads, navigation, and sidebars. Pulpie removes this noise and returns a clean document you can use in pre-training or as context.
Previously, cleaning 1 billion pages with the best extractors cost $159,000. Pulpie brings that down to $7,900. A 20x decrease.
English
We launched @amikadev: the open source operating system for AI agents to automate coding and prove the code correct
Amika runs fleets coding agents in cloud VMs to automate flows like:
- fix production errors, with a test proving the fix
- ship user feature requests from Slack
English