Remote Customer Reliability Engineer for AI Compute

OpenKrill; United States; Remote; USD 120,000 - 180,000; Full time

Benefits offered by this job

Equity

Healthcare coverage

Unlimited PTO

Competitive compensation

Job summary

Andromeda is seeking a Customer Reliability Engineer to support AI compute workloads remotely across APAC. You will triage incidents, debug Linux and Kubernetes issues, and work with providers to drive fast resolution.

The role requires strong scripting, documentation, and a willingness to be on a 24/7 on-call rotation. You’ll join a team focused on reliability for large GPU clusters, build tooling, and create runbooks that reduce future incidents.

Qualifications

• Real Linux troubleshooting ability from the command line.
• Working knowledge of Kubernetes: pods, nodes, deployments, scheduling, and debugging.
• Ability to write a script in Python or Bash to automate repetitive tasks.
• Strong writing to explain technical problems clearly to customers.
• Willingness to join a 24/7 on-call rotation.

Responsibilities

• Triage and fix customer issues end-to-end: reproduce, diagnose, fix or elevate, close the loop.
• Debug at the Linux layer: processes, networking, storage, kernel logs, resource contention, systemd, journald.
• Investigate Kubernetes issues: pending pods, crash loops, node conditions, scheduling.
• Address GPU-related failures: drivers, device-plugin, XID errors, node cordon/drain.
• Escalate when needed with evidence and clear communication.
• Build and maintain runbooks to reduce future incidents.
• Write Python or Bash to automate checks and daily tasks.
• Develop monitoring and dashboards for health and alerts.
• Keep customer docs up to date as you go.

Skills

Linux troubleshooting

Kubernetes

Python or Bash scripting

Effective written communication

On-call readiness

Tools

Linux

Kubernetes

Job description

Andromeda is seeking a Customer Reliability Engineer to support AI compute workloads remotely across APAC. You will triage incidents, debug Linux and Kubernetes issues, and work with providers to drive fast resolution.

The role requires strong scripting, documentation, and a willingness to be on a 24/7 on-call rotation. You’ll join a team focused on reliability for large GPU clusters, build tooling, and create runbooks that reduce future incidents.