Senior HPC DevOps Engineer - Remote, GPU Cloud & MLOps

EPAM Systems; Polska; Hybrid; PLN 180,000 - 280,000; Full time

Benefits offered by this job

Health insurance

Multisport

Shopping vouchers

Employee Stock Purchase Plan discounts

Relocation opportunities

English classes

Job summary

EPAM Systems in Poland seeks a Senior HPC DevOps Engineer to scale and automate our HPC and MLOps platform, guiding scientists to run self-service workloads with minimal friction and clear governance.

You will optimize capacity, manage AWS GPU instances, deploy containerized environments, profile and tune workloads, and collaborate across teams to improve observability, documentation, and performance for large-scale AI pipelines.

Qualifications

• 5+ years of experience in HPC or DevOps engineering roles.
• Knowledge of MPI, OpenMP, and multi-node GPU communication protocols such as NCCL and GPUDirect.
• Experience managing AWS GPU instance families including P-series, G-series, and Tranium/Inferentia.
• Hands-on mastery of AWS Capacity Blocks for ML, On-Demand Capacity Reservations (ODCRs), and Service Quota management.
• Containerized environments using Apptainer/Singularity, Docker, or Enroot.
• Understanding of I/O bottlenecks with distributed file systems like Lustre, GPFS, BeeGFS, or AWS FSx for Lustre.
• Profiling applications with NVIDIA Nsight or similar tools to locate memory and compute bottlenecks.
• Experience deploying HPC workloads on cloud using EFA, ParallelCluster, and parallel storage (FSx Lustre).
• Proficiency in English at a B2+ level.

Responsibilities

• Guide scientists and data teams in navigating the platform UI for self-service workloads.
• Advise users on capacity planning, quotas, and cost for heavy workloads.
• Maintain automated pipelines for infrastructure provisioning and platform deployments.
• Resolve job scheduling failures, cluster bottlenecks, and quota issues.
• Collaborate to improve documentation and developer experience.
• Monitor GPU utilization with tools like Prometheus and CloudWatch.
• Manage AWS GPU instance families and capacity reservations for ML workloads.
• Deploy and scale HPC workloads on cloud with parallel storage and networking.
• Deploy containerized HPC environments tuned for GPU pass-through.
• Collaborate with engineering teams to optimize performance and reliability.

Skills

MPI

OpenMP

NCCL

GPUDirect

English proficiency

Tools

Docker

Apptainer/Singularity

Enroot

NVIDIA Nsight

Prometheus

CloudWatch

AWS Capacity Blocks

ODCRs

ParallelCluster

FSx for Lustre

Job description

EPAM Systems in Poland seeks a Senior HPC DevOps Engineer to scale and automate our HPC and MLOps platform, guiding scientists to run self-service workloads with minimal friction and clear governance.

You will optimize capacity, manage AWS GPU instances, deploy containerized environments, profile and tune workloads, and collaborate across teams to improve observability, documentation, and performance for large-scale AI pipelines.