Remote GPU Cloud Infrastructure Specialist

The POD Network; Polska; Hybrid; PLN 461,000 - 614,000; Full time

Benefits offered by this job

Equity

Medical, dental & vision plans

Flexible PTO

Remote-friendly culture

Home Office stipend

Job summary

Runpod, the AI Developer Cloud, seeks a Datacenter Infrastructure Specialist to own hardware lifecycle and fleet health. You will interface with hardware partners and internal engineering, ensuring high uptime for AI workloads at scale.

You will work with RDMA networks, NVIDIA software, and Linux-based automation to drive reliability, performance, and incident response. Remote-first, globally distributed team.

Qualifications

• 3–5 years in infrastructure operations, reliability, or datacenter engineering.
• Strong datacenter networking; exposure to RDMA/InfiniBand/RoCE preferred.
• Hands-on NVIDIA software stack and multi-node performance tuning.
• Solid Linux admin skills and Docker experience.
• Effective written and verbal communication; able to explain complex issues.
• May participate in on-call rotation as Fleet scales.

Responsibilities

• Hardware validation and benchmarking for new deployments.
• Monitor fleet uptime and enforce SLA through data-driven analysis.
• AI-driven operations with automation and runbooks.
• Coordinate incident communications and provide actionable updates.
• Support growth of infrastructure partners and internal teams.

Skills

Infrastructure ops

Datacenter networking

RDMA/InfiniBand/RoCE

NVIDIA software stack

Linux administration

Docker

Clear communication

On-call readiness

Strategic problem solving

Tools

NVIDIA Software Stack

Docker

Grafana/Prometheus/Datadog

Job description

Runpod, the AI Developer Cloud, seeks a Datacenter Infrastructure Specialist to own hardware lifecycle and fleet health. You will interface with hardware partners and internal engineering, ensuring high uptime for AI workloads at scale.

You will work with RDMA networks, NVIDIA software, and Linux-based automation to drive reliability, performance, and incident response. Remote-first, globally distributed team.