Senior GPU Engineer

 Senior GPU Engineer – Decentralised Compute Infrastructure – Global Remote

Plexus are working with one of the most exciting teams in the decentralised computing space. With a focus on enhanced scalability, efficiency, cost and security. 

With a recent raise of 8figures, they are looking to onboard a hands on Senior GPU Engineer into the team. 

Responsibilities

• Design and manage multi-tenant GPU clusters using Kubernetes, Slurm, or similar platforms.
• Develop schedulers and resource-sharing tools to optimize GPU utilization and efficiency.
• Build autoscaling systems for training and inference workloads (Ray, Run:AI, Volcano, KubeFlow)
• Implement monitoring and observability for GPUs, network, and job performance (Prometheus, Grafana, OpenTelemetry).
• Profile and optimize compute throughput and cost efficiency (NCCL, CUDA, ROCm, GPUDirect, RDMA, InfiniBand).
• Collaborate on high-throughput I/O and data pipelines (Lustre, Ceph, S3, NVMeoF, Alluxio).

Requirements

• 2+ years managing distributed GPU infrastructure in production.
• Strong experience with Kubernetes or Slurm, and Linux systems.
• Skilled in Python/Go/C++, automation, and infrastructure-as-code (Terraform, Helm).
• Familiar with CUDA/NCCL/ROCm, Ray/Run:AI/Volcano, and high-speed networking (InfiniBand, RoCE).
• Knowledge of AI storage systems (Lustre, Ceph, S3) and performance optimization.
• Excellent collaboration and communication skills across technical teams.

Offer/Benefits

• Up to $220k base 
• Fully remote, work-from-anywhere 

Sound like you, or someone you know? Please apply