[AI Research Div.] AI Infra Engineer Lead (8년 이상)
Salary not listed • Seoul; Seoul, South Korea • Full-time
Posted 10d ago
Remote Work Policy
On-site
Employment Type
Full-time
Categories
AI Infrastructure Engineer
About the job
KRAFTON AI Research is seeking a Lead AI Infrastructure Engineer to build and operate the GPU infrastructure necessary for model development, along with AI platforms based on Kubernetes and Slurm. This role involves managing GPU clusters, model training and experimentation environments, ML pipelines, infrastructure automation, and observability systems to ensure AI workloads for game development and services run stably and efficiently. The position is for a hands-on senior engineer who will operate and enhance existing large-scale GPU infrastructure, design future infrastructure expansions, and ensure stable and efficient utilization by R&D and service teams. You will also contribute to technical reviews and proposals for future computing infrastructure scaling and KRAFTON's GPU and computing operational strategy.
Responsibilities
- Operate and enhance large-scale GPU infrastructure for stable and efficient use by R&D/service organizations.
- Stabilize operations, improve performance, optimize resource utilization, and automate operations for large-scale Blackwell Ultra-based GPU infrastructure.
- Design, build, and operate Kubernetes-based ML/GPU platforms, including scheduling, multi-tenancy, workload isolation, quotas, observability, and failure response.
- Develop and implement operational strategies to improve GPU utilization, latency, throughput, and cost-efficiency based on training/inference workload characteristics.
- Conduct technical reviews and propose future computing infrastructure expansion directions based on GPU capacity planning and utilization analysis.
- Enhance GPU-based reproducible operating systems and coordinate requirements from various teams from a common platform perspective.
Requirements
- Experience designing, building, and operating large-scale GPU clusters or Kubernetes/Slurm-based AI/ML infrastructure.
- Understanding of computing infrastructure including GPU servers, networks, and storage, with experience building stable operating environments.
- Experience applying GPU resource management, scheduling, multi-tenancy, workload isolation, and quotas in an operating environment.
- Experience optimizing infrastructure by analyzing GPU utilization, performance, stability, and cost.
- Experience analyzing root causes of failures and performance issues from a system perspective and implementing structural improvements.
- Experience collaborating with R&D/service organizations to define infrastructure requirements and propose/execute expansion and operational directions.
- No disqualifications for overseas business trips.