Head of GPU Compute Production Engineering

Fluidstack; San Francisco (CA); On-site; USD 269,000 - 335,000; Full time

Job summary

Fluidstack is seeking a leader for the Data Center Operations Team to steward tens of thousands of GPUs and ensure high availability for customer workloads. The role emphasizes automation, end-to-end ownership, and fast, reliable operations at scale in a frontier compute environment.

Ideal candidates have led SRE or production engineering teams, shipped automated remediation, and can articulate how to improve availability metrics. Bonus: Kubernetes, Slurm, and HPC familiarity.

Qualifications

• Led SRE or production engineering teams running large fleets.
• Able to explain how to move an availability number.
• Shipped automated remediation that retired a runbook.
• Has hired and grown engineers; team growth track record.

Responsibilities

• Lead the compute production engineering team keeping tens of thousands of GPUs serving customers.
• Own fleet availability for compute: define the SLOs, build the tooling, and move the number.
• Build automation for the node lifecycle, provisioning, health checks, remediation, return to service, without human touch.
• Set the on-call and escalation model that keeps response sharp without burning the team out.

Skills

Led SRE teams

Production engineering

Automated remediation

Hiring and growing engineers

GPU/HPC familiarity

Tools

Kubernetes

Slurm

GPU HPC fleets

Job description

Fluidstack is seeking a leader for the Data Center Operations Team to steward tens of thousands of GPUs and ensure high availability for customer workloads. The role emphasizes automation, end-to-end ownership, and fast, reliable operations at scale in a frontier compute environment.

Ideal candidates have led SRE or production engineering teams, shipped automated remediation, and can articulate how to improve availability metrics. Bonus: Kubernetes, Slurm, and HPC familiarity.