StarSling, Inc.
StarSling is an agent-native CI platform for GitHub Actions. It provides fast managed runners and AI agents that analyze workflows and continuously open optimization pull requests so software teams spend less time waiting on CI and more time shipping.
At a Glance
- AI product companies
- Developer-tool companies
- Software engineering teams using GitHub Actions
- Platform and infrastructure teams
- +1 more
AI Tools by StarSling, Inc.
(1)StarSling
AI Runners for GitHub Actions
Discussions
No discussions yet
Be the first to start a discussion about StarSling, Inc.
Latest News
Announcing the CI Challenge: save 10% on CI or get $1,000
Introducing Disks: files ready 10.5x faster
Introducing CI Playbooks: 250+ GitHub Actions Lessons from Software Factories
OpenAI's CI playbook for GitHub Actions
Products & Services
Managed Ubuntu GitHub Actions CPU runners on 5th Gen AMD EPYC hardware, with drop-in runs-on labels, unlimited concurrency, per-second billing for new paid subscriptions, and agents that open workflow-optimization pull requests.
Dedicated NVIDIA GPU runners for GPU-dependent tests, evaluations and CUDA workloads, currently offering RTX PRO 6000, RTX 5090 and RTX 4090 configurations through GitHub Actions; H100 is listed as coming soon.
Repository-defined code-review agents that run as GitHub Actions jobs using the team's model, repository skills, scripts and reviewer responsibilities; customers pay StarSling for compute and their model provider for tokens.
Persistent filesystems mounted into GitHub Actions jobs to make dependencies and other files available faster.
Market Position
StarSling positions itself as AI-native CI rather than only a faster runner: it combines managed GitHub Actions-compatible CPU/GPU compute with agents that continuously improve workflows through reviewable pull requests. Its own comparison pages name GitHub-hosted runners, Blacksmith and Depot as alternatives; StarSling differentiates on agent-generated optimization PRs, Adaptive Compute, CI skills and per-second billing.
Leadership
Founders
Yonas Beshawred
Co-founder and CEO. Previously founded StackShare, a developer community with more than 1.5 million developers; StackShare was acquired by FOSSA in 2024.
Daniel Worku
Co-founder and CTO. Led the 12-person engineering team behind Netflix Console, an internal developer portal serving more than 3,000 engineers, managers and leaders; previously worked on Facebook's AI Camera team.
Executive Team
Yonas Beshawred
Co-founder & CEO
Founded StackShare, a developer community with more than 1.5 million developers; StackShare was acquired by FOSSA in 2024.
Daniel Worku
Co-founder & CTO
Led the engineering team behind Netflix Console for more than 3,000 engineers and previously worked on Facebook's AI Camera team.
Founding Story
Daniel Worku and Yonas Beshawred started StarSling in 2025 because production DevOps tooling was falling behind local coding agents. While building agents for that broader problem, they encountered CI as a bottleneck themselves and built the runners and optimization agents they wanted to use.
Business Model
Revenue Model
Usage-based compute billing for runner time. New paid subscriptions bill by the second with no seat fees and queue time excluded; Review Runners use the same runner billing while model-token charges are paid directly to the customer's model provider.
Pricing Tiers
Ubuntu 26.04 or 24.04; per-second rate listed as $0.00006667.
Default configuration; per-second rate listed as $0.00013333.
Usage-based compute.
Includes 4 vCPU, 16 GB RAM and 100 GB disk; GPU access is opened for an organization on request.
Dedicated GPU runner; access by request.
Dedicated GPU runner; access by request.
For new accounts; applies to the first month.
Target Markets
- AI product companies
- Developer-tool companies
- Software engineering teams using GitHub Actions
- Platform and infrastructure teams
- Teams running GPU tests or model evaluations
- Speeding up build, test and end-to-end workflows in GitHub Actions
- Reducing CI compute cost and queue time for high-volume repositories
- Automatically improving caching, dependency installation, parallelization and test sharding
- GPU-dependent tests and model evaluations
- Running repository-defined AI code reviews on every pull request
- Helping platform teams manage growing CI usage and compute costs
- Better Auth
- Mastra
- Partcl
- Bespoke Labs