Senior MLOps Engineer

Role overview Keep machine learning models dependable and observable in production on client GPU infrastructure. You’ll build and operate the serving, scheduling, and lifecycle systems behind private language model and retrieval-augmented generation platforms used by enterprise clients in South Korea.

Responsibilities - Deploy and operate models on GPU clusters, with attention to reliable serving and resource scheduling. - Build pipelines for data preparation, model training, and evaluation. - Manage model versions and support controlled releases. - Monitor production latency, infrastructure cost, and model quality in on-premise environments. - Work closely with a small, experienced team on client-facing platform deployments.

Requirements - At least five years of experience in MLOps or ML platform engineering. - Production experience with model serving and GPU scheduling, using tools such as vLLM, Triton, or Kubernetes. - Experience building data, training, and evaluation pipelines and managing model versions. - Ability to monitor and troubleshoot latency, cost, and quality for on-premise deployments. - Valid residency in South Korea; a residence card or ID card is required with the application. - Working English; Korean is helpful.