We are looking for a Lead, Middle, or Senior DevOps Engineer to join a research infrastructure team building an on-demand GPU platform for advanced compute workflows. The role focuses on enabling secure, scalable, and user-friendly access to high-performance GPU resources through automation, scheduling, and modern platform tooling.
Locations: Serbia, Georgia, Armenia, Kazakhstan, Poland, Croatia, Portugal, Egypt.
Tasks
- Strong hands-on experience with Kubernetes and platform orchestration;
- Solid understanding of scheduling, reservation, or namespace-based resource management systems;
- Experience with GPU infrastructure, virtualization, slicing, or containerized workstation environments;
- Strong scripting and automation skills;
- Practical Azure experience and familiarity with secure infrastructure operations.
Requirements
Responsibilities
- Build and improve an on-demand GPU workstation platform with lightweight containerization or virtualization;
- Implement scheduling, reservation, registration, image management, storage mounting, SSH with SSO, and developer-friendly access flows;
- Automate cluster namespace configuration across CPU, GPU, memory, and storage allocations;
- Support hierarchical capacity allocation models with RBAC-based administration;
- Automate storage import, export, and archival workflows as allocations change;
- Build monitoring, alerts, and automated incident ticket creation for large-scale cluster environments;
- Improve integrations between source control, CI/CD, package distribution, and GPU-connected development workflows;
- Contribute automation, scripts, and agentic tooling that improve infrastructure and day-to-day research workflows.
Nice to Have:
- Experience with Prometheus, Grafana, incident automation, or on-call paging workflows;
- Experience with developer platforms, devcontainers, or remote development tooling such as VS Code integrations;
- Exposure to AI-assisted monitoring, trend analysis, or agentic infrastructure tooling.
Engagement Type
Location / Timezone
- Remote work from Serbia, Georgia, Armenia, Kazakhstan, Poland, Croatia, Portugal, Egypt.
- European working hours.
- Occasionally available for meetings up to 10:00 AM PST (US overlap).