[AI Research Div.] ML Serving Engineer (5년 이상)

Posted about a month ago

or apply directly on KRAFTON's site. We never take the application ourselves.

Is this posting real?

This role has been open: 30 days
Reposted: No
Salary listed: No
Ghost-job risk at KRAFTON: high
Hiring momentum: 92 roles opened in the last 90 days
Last confirmed on the employer's board: 2026-09-17

Measured from postings appearing on and disappearing from KRAFTON's own greenhouse board since 2026-08-03. Full hiring picture for KRAFTON.

About this role

The ML Serving Engineer will design and implement high-performance ML serving architectures for game applications, ensuring stable and efficient operation of AI models in production environments. Responsibilities include building real-time serving infrastructure for large language models (LLMs), optimizing performance, and establishing CI/CD pipelines for continuous deployment and monitoring.

Our read on this posting3.0out of 5

benefits: 1/5
freshness: 4/5
career value: 5/5
role clarity: 5/5
pay transparency: 0/5

Scored from the posting itself — how clearly the role is described, how much it says about pay and benefits, and how recently it was listed. Not a judgement of KRAFTON as an employer.

What you need

5+ years of experience in designing, building, and operating AI/LLM model serving systems in production environments
Experience applying high-performance LLM inference engines like vLLM or TensorRT-LLM for system optimization
Knowledge of memory and computation efficiency techniques such as KV Cache Offloading/Loading and quantization
Strong analytical skills to address latency and performance issues across inference engines and GPU hardware
Experience building and enhancing monitoring and observability systems for serving stability and CI/CD automation
Ability to travel internationally without restrictions

Nice to have

Experience in making architectural decisions and understanding trade-offs in performance, cost, and stability
Contributions to open-source serving frameworks like vLLM or LMCache
Deep understanding of high-performance LLM inference stacks and experience in custom kernel development and CUDA acceleration

Worth weighing

No salary or specific benefits mentioned
The role requires a significant level of expertise and experience, which may limit applicants
The position involves a probation period of 5 months with no changes in employment type or salary during this time

Summarised from KRAFTON's posting. Read the full original.

Listed by KRAFTON on their greenhouse job board, last confirmed open on 2026-09-17. PitchMeAI is not the employer.

More roles at KRAFTON

All 61 open roles at KRAFTON →