In short
Administers Fabrinet's on-premise Linux GPU server and cluster platform that hosts AI services for its manufacturing operations — handling OS/NVIDIA driver patching, container platform and AI model deployments, monitoring, backups, and access control. Core stack: Linux, Docker/Kubernetes, CUDA model serving (NIM, vLLM, Ollama), Bash/Python scripting.
About the role
Job Purpose :
Maintain a secure, reliable on-premise AI server and cluster platform for engineers and AI developers. Manage system patching, application and approved model deployment, monitoring, backup and recovery to support manufacturing AI services and protect company intellectual property.
Key Responsibilities :
1. Administer Linux servers, GPU nodes, storage and cluster resources; coordinate hardware, network and facility issues with IT and vendors.
2. Apply tested OS, firmware, security, NVIDIA driver and runtime updates. Verify compatibility, schedule maintenance and maintain rollback plans.
3. Install, configure and maintain container platforms, AI applications, model-serving services and supporting databases using controlled deployment procedures.
4. Deploy and version approved AI models and dependencies; record source, license approval, configuration and resource needs. Verify API availability and support rollback.
5. Monitor service health, GPU memory and utilization, storage, logs and API latency. Troubleshoot incidents and track corrective actions.
6. Manage role-based access, service accounts, secrets, certificates and audit logs. Enforce business-unit separation and approved software and model intake.
7. Maintain backups of configurations, application data and essential model artifacts; test restoration and document recovery procedures.
8. Automate routine tasks, maintain inventories and operating guides, support AI developers and users, and report service availability, patch status and capacity risks.
Job Qualifications :
Education: Bachelor’s degree in Computer Science, IT, Computer Engineering or a related field, or equivalent practical experience.
Experience: At least 1-5 years in systems administration, including hands-on Linux and production container operations. GPU or AI platform experience is preferred.
Required Skills: Linux troubleshooting; Docker and Kubernetes administration where deployed; networking and storage fundamentals; patching, backup, monitoring and access control; Bash or Python scripting; Git and change management.
Preferred Skills: NVIDIA GPU drivers, CUDA compatibility and model serving such as NVIDIA NIM, vLLM or Ollama; Helm, CI/CD, Prometheus/Grafana and object storage.
Communication: Able to work with engineers, developers, IT security and vendors; read technical English and produce clear operational documentation.
Job Competencies :
Systematic troubleshooting; security and service ownership; careful change planning; clear communication; teamwork; accurate documentation; and continuous learning.