ciy405x@gmail.com / (+82) 10-9504-8680 / Seoul, South Korea / LinkedIn

Machine Learning Engineer with 4 years of experience since 2021.
Focus: 3D Deep Learning, Generative Models, Cross-Platform Systems, Performance Optimization.

Currently building production-ready 3D content generation pipelines based on diffusion and language models.
Main interest: making 3D ML models run faster, cheaper, and more useful for games, graphics, and visual applications.

• World’s first in-game 3D generation system for consumer release
  → Deployed in inZOI, enabling players to generate 3D furniture, props, and accessories from a single photo within 15 seconds
  → Ported across architectures including Intel, ARM, and Apple Silicon
  → Became the most beloved AI feature, widely shared by users (over 1 million creations):
  • YouTube fan showcase
  • inZOI Canvas fan uploads
  • Reddit fan thread
• Real-time text-to-building generation using small language models (SLM)
• ACM TOG (proc. of SIGGRAPH) paper
• Granted US patent

• Languages: Python, C++, JavaScript
• Frameworks: PyTorch, TensorFlow, ONNX
• Graphics: CUDA, OpenGL, WebGL
• Infra/Tools: Unreal Engine, Blender

• Current: ML Engineer at KRAFTON (the PUBG company), deploying AI-driven content creation pipelines
• Previous: ML Engineer at SPACEWALK, optimizing architectural design via reinforcement learning (RL)
• Education:
  • MS in Computer Science, KAIST
    • Thesis on Monte Carlo rendering denoising
    • GPA 4.25 / 4.30
  • BS in Computer Science and Mathematics, KAIST
    • GPA 4.00 / 4.30
    • Double major
    • Dean's List (Top 3%) for 3 consecutive semesters
  • Most enjoyably immersed courses: Computer Graphics, Operating Systems, Real/Complex/Numerical Analysis, Linear Algebra, Matrix Computation, 3D Machine Learning, and GPU Programming
  • Advised and mentored 50 undergraduate students as a counseling assistant