Customer Reliability Engineer

Fluidstack•San Francsisco, CA

9d•$204,000 - $284,000

About The Position

Fluidstack is building civilization-scale infrastructure for AI, aiming to deliver 10 to 100s of GWs of compute faster than anyone else. This involves rethinking every layer of the stack, from acquiring power to designing, building, and operating data centers. The company operates with a philosophy of extreme ownership, full autonomy, velocity, first principles thinking, and a deep passion for the problem space. The Data Center Operations Team is responsible for operating at the scale of a nation, managing sites that come online in pieces while ensuring live operations remain flawless. They are tasked with writing the playbook for operating at this unprecedented speed and scale, as no prior operations organization has done so. The Customer Reliability Engineer role focuses on owning the reliability for named customer workloads, including their clusters, SLAs, and escalations. This involves debugging across the full stack (hardware, fabric, scheduler) when training runs degrade, managing customer-facing incident communication with technical depth and transparency, and transforming recurring customer pain points into engineering fixes with the production teams.

Requirements

Supported large-scale compute customers (HPC, cloud, or AI labs) at a technical level.
Debug distributed systems methodically across layers you don't own.
Written incident updates customers trusted more after reading.
Push internal teams to fix causes, not symptoms, and follow up until they do.

Nice To Haves

GPU training workloads.
InfiniBand or RoCE.
Slurm or Kubernetes.
NCCL debugging.

Responsibilities

Own reliability for named customer workloads: their clusters, their SLAs, their escalations.
Debug across the full stack, hardware to fabric to scheduler, when a training run degrades.
Run customer-facing incident communication with technical depth and no spin.
Turn recurring customer pain into engineering fixes with the production teams.