- Experience
- Any
- Salary
- SGD 200,000 – SGD 400,000 / year
- Openings
- 1
- Posted
- 2 weeks ago
- Work mode
- Work from home
- Eligibility
- Candidates with strong CI/CD, infrastructure, and automation experience who can work from Singapore and contribute to large-scale developer tooling are suited for this role. Visa sponsorship may be available case by case.
- Resume
- Required to apply
Job description
About the company
Inferact is building vLLM to become the leading AI inference engine, with the goal of making inference faster and more affordable while helping advance AI progress. The company was started by the founders and core maintainers of vLLM, bringing together deep expertise in models and hardware.
About the role
vLLM is expanding quickly, and that growth puts increasing pressure on the CI system. As the codebase, hardware variety, and contributor base scale, this role focuses on keeping engineering velocity high by ensuring the build and test infrastructure remains dependable and fast.
What you'll do
- Own and expand the compute infrastructure that supports CI, releases, performance benchmarking, and accuracy evaluation for the vLLM project.
- Support workloads across a broad set of accelerators and systems, including H100/H200, (G)B200/300, AMD MI325/355X, TPU, Intel Gaudi, and similar platforms.
- Reduce the time it takes for CI to surface meaningful signals, moving from hours down to minutes where possible.
- Improve test coverage so that all major areas of the vLLM codebase remain well validated.
- Help keep releases stable, reliable, and ready for production use.
- Create tools and internal systems that allow more than 3,000 vLLM contributors to work efficiently.
Minimum qualifications
- Hands-on experience with Docker, Kubernetes, and container-based build or test environments.
- Experience designing and building CI/CD pipelines from the ground up using GitHub Actions, Buildkite, or similar platforms.
- Working knowledge of CI patterns and practices such as compute orchestration, flaky test handling, dependency and environment management, caching, remote execution, and test target selection.
- Comfort using Python, Bash, Go, or comparable languages for automation and tooling.
- Strong understanding of Linux fundamentals, plus practical knowledge of security, networking, storage, and package management.
Bonus experience
- Experience setting up infrastructure for ML, inference, CUDA, ROCm, or other accelerator-intensive workloads.
- Prior work operating Buildkite at scale, including agents, queues, dynamic pipelines, test sharding, caching, and artifact handling.
- Background managing Kubernetes clusters for CI, batch processing, test execution, or internal developer platforms.
- Experience supporting CI/CD for a large open-source project.
- Ability to build dashboards, alerts, runbooks, or other observability tooling for CI systems.
Logistics
This position is based in Singapore. Compensation is expected to be between S$200,000 and S$400,000 per year, plus equity, depending on background, skills, and experience. Visa sponsorship may be provided on a case-by-case basis. The benefits package includes medical, dental, and vision coverage.
Compensation
The stated annual compensation range is SGD 200K to SGD 400K, with equity in addition to base pay.