Senior ML Infrastructure Engineer

Responsibilities

• Operate and optimize B300 125-node GPU infrastructure for stability and efficiency
• Design and manage Kubernetes-based ML/GPU platforms including scheduling, multi-tenancy, and observability
• Develop operational strategies to improve GPU utilization, latency, and throughput
• Coordinate requirements from research and development teams to build a common ML platform

Requirements

• Experience designing and operating large-scale GPU clusters for AI/ML workloads
• Experience improving ML/GPU platform scheduling, workload isolation, and fault tolerance
• Ability to analyze system-wide performance issues and implement structural improvements
• Experience collaborating with R&D teams to define infrastructure requirements
• Proficiency in using AI tools (LLMs, code assistants) to enhance operational productivity
• Ability to travel overseas if required

Preferred Qualifications

• Experience with next-generation GPU architectures (B300, H100, H200, GB200)
• Expertise in GPU networking optimization (NCCL, RDMA, RoCE, InfiniBand)
• Experience with distributed storage (Ceph, MinIO) optimized for AI workloads
• Proficiency with GPU management tools (NVIDIA GPU Operator, DCGM, MIG/MPS, Slurm, Volcano)
• Proven track record of improving training throughput and GPU utilization in distributed environments

About the Company

KRAFTON is a global gaming company dedicated to creating unforgettable worlds for fans through bold imagination and breakthrough technology.