Juntak Noh{width=120px}

I’m Juntak Noh, an applied AI engineer focused on building and owning production-grade machine learning systems.

I have 5+ years of experience designing and deploying real-time AI systems across computer vision, statistical modeling, and LLM-based reasoning. My work spans from low-latency C++ inference pipelines integrated into live game environments to Python-based ML systems that translate complex data into actionable feedback.

I’m particularly interested in:

  • Failure modes of ML systems in real-world settings
  • Bridging research insights and production constraints
  • LLM systems beyond prompt engineering

This site is a public record of how I think and work, including:

  • Paper reviews (assumptions → failure modes → what I’d change)
  • Applied research experiments
  • Lessons from building ML systems at scale

I hold an M.S. in AI from Korea University and a B.S. from DGIST (cum laude). I’m fluent in English and have worked in international research and engineering environments.

Outside of AI, I’m an active pianist (International Chopin Competition finalist) and a professional poker player, both of which strongly shape how I approach discipline, uncertainty, and long-term decision making.

📄 Full CV: Download PDF
📬 Contact: [email protected]