Reliability Engineer for Large-Scale AI Workloads

Fluidstack; New York (NY); Hybrid; USD 180,000 - 260,000; Full time

Job summary

Fluidstack is seeking a Production Engineer to own reliability for high-demand customer workloads and SLA management. You will debug across the full stack, from hardware to scheduler, ensuring live systems stay online during AI training runs.

You will draft precise incident updates, drive root-cause analyses, and push the engineering teams to fix underlying causes rather than symptoms. GPU workloads, InfiniBand, Slurm, and Kubernetes experience are a plus.

Qualifications

• Experience supporting large-scale compute customers (HPC, AI labs, cloud).
• Ability to debug distributed systems across multiple layers.
• Proven track record communicating incidents with technical depth and timely updates.
• Experience driving root-cause analysis and follow-up actions with engineering teams.
• Bonus: GPU training workloads, Slurm, NCCL, InfiniBand.

Responsibilities

• Own reliability for named customer workloads and SLAs.
• Debug across hardware to fabric to scheduler during training runs.
• Manage customer-facing incident updates with clear, technical communication.
• Turn recurring customer pain into engineering fixes with production teams.

Skills

Distributed systems debugging

Incident communication

Customer-focused support

HPC/AI compute experience

Tools

Kubernetes

Slurm

NCCL

InfiniBand

Job description

Fluidstack is seeking a Production Engineer to own reliability for high-demand customer workloads and SLA management. You will debug across the full stack, from hardware to scheduler, ensuring live systems stay online during AI training runs.

You will draft precise incident updates, drive root-cause analyses, and push the engineering teams to fix underlying causes rather than symptoms. GPU workloads, InfiniBand, Slurm, and Kubernetes experience are a plus.