[AI Frontier Div.] AI Inference Platform & Operations Engineer (5년 이상)
Core
Building and operating a shared AI inference platform and control plane for large-scale LLM/ML model serving, ensuring stability, low latency, and cost efficiency for game and other products.
Role type
Senior IC AI Inference Platform & Operations Engineer
Builds
Kubernetes-based GPU clusters, serving engines (vLLM, SGLang), orchestration layers, and automated model onboarding/deployment pipelines.
Domain
AI/ML Infrastructure, Cloud Native, Game Tech
Deliverable
production ML models
Required skills
Kubernetes, GPU resource management, SRE practices, observability (metrics/logs/traces), IaC/GitOps/CI/CD, model serving optimization, system design
Preferred skills
Large-scale low-latency service operations, capacity planning, vLLM/SGLang/TensorRT-LLM/Triton Inference Server, continuous batching/KV cache management, multi-node GPU optimization, research-to-production integration
Technologies
Kubernetes, vLLM, SGLang, TensorRT-LLM, Triton Inference Server, GPU clusters
Responsibilities
Operate and improve serving platforms and control planes; automate safe model onboarding and deployment; establish SLOs and monitor reliability; manage GPU resources and unit cost; implement serving logic and optimize serving systems; validate and apply research optimizations to production.
Seniority
Senior, hands-on IC