Zyphra
Machine Learning Systems Administrator / DevOps Engineer: Training Clusters & Inference Cloud
San Francisco · On-site · Full-time · Posted Oct 15, 2026
- Visa sponsorship
- Relocation and immigration support on a case-by-case basis.
About the job
Maintains and builds infrastructure for model training clusters and production inference services across heterogeneous compute.
What you'll do
- Administer and automate Linux clusters, including access control, monitoring, hardware validation, and support for training workloads.
- Deploy, operate, and scale inference services; manage orchestration, CI/CD, observability, and service reliability.
- Plan capacity and utilization, secure production systems, and support developer workflows.
What they ask for
- Experience with Linux system administration, user and access management, and automation.
- Demonstrated scripting expertise for system tooling and automation (e.g., bash, Python).
- Experience with Docker and scheduling/orchestration systems such as Slurm or Kubernetes.
- Experience operating production services, including monitoring, alerting, deployment pipelines, and incident response.
Nice to have
- Infrastructure-as-code experience (e.g., Ansible, Terraform) and prior ML/AI infrastructure work, including GPU management.
- Experience with LLM serving stacks, multiple accelerator platforms, high-performance networking, distributed storage, cloud platforms, or Apptainer.
Tools
Linux, bash, Python, Docker, Slurm, Kubernetes, Prometheus, Grafana
AI focus
Supports large-scale model training and operates production inference services, including LLM serving stacks as a preferred qualification.
Model-written summary · the original posting is authoritative