GPU Cluster Infrastructure Engineer

Location: On-site in Metzingen / Riederich, Germany

Company

hirify.global develops robotics products supported by advanced software engineering and machine learning infrastructure.

What you will do

• Operate and continuously evolve large-scale AWS HyperPod GPU clusters using Slurm and EKS orchestration models.
• Design cluster-stability mechanisms, including node-failure detection, automated job recovery, checkpoint coordination, and fault-tolerant multi-node training.
• Optimize GPU utilization across compute, memory, EFA networking, and storage throughput.
• Build self-service tooling for ML researchers and engineers to launch, monitor, and manage training workloads.
• Define workload-priority, capacity, and cost-management strategies across pretraining, fine-tuning, and customer workloads.
• Collaborate with AWS HyperPod teams and internal ML, product, finance, and cloud vendor stakeholders.

Requirements

• 5+ years of infrastructure or systems engineering experience, with a strong focus on GPU cluster or HPC operations.
• Hands-on experience with AWS HyperPod and AWS GPU instances.
• Strong knowledge of Slurm and Kubernetes and the trade-offs between them for large-scale GPU workloads.
• Practical understanding of distributed training, including throughput analysis and debugging.
• Experience building self-service tooling and operational documentation for technical users.
• Strong English communication skills required; German is a plus.

Culture & Benefits

• Infrastructure changes and cluster configurations are managed through Infrastructure as Code.
• Work directly with AWS product and solutions engineering teams to report issues and influence the HyperPod roadmap.
• Create onboarding documentation, training materials, and internal workshops for ML infrastructure users.
• Manage Spot interruption handling, capacity reservations, cost attribution, Reserved Instances, Savings Plans, and AWS commitment negotiations.