Hi, I’m Wookje Han, an AI Developer Technology Engineer at NVIDIA. My work focuses on (1) e2e performance optimization for LLM, video, and recommender systems and (2) kernel-level optimization across pretraining and inference. Some of my work includes designing and implementing SOTA attention and GEMM kernels on Hopper, Blackwell, and Rubin arch.

Previously, I received my M.S. in Computer Science from Columbia University and my B.S. in Computer Science from Seoul National University (Summa Cum Laude).

Feel free to reach me at wookjeh@gmail.com!

---

Writing & Contributions

---

• NVIDIA Platform Delivers Lowest Token Cost Enabled by Extreme Co-Design (Apr 2026)

---

Publications

---

* indicates equal contribution

Accepted Papers

PreWoMe: Exploiting Presuppositions as Working Memory for Long Form Question Answering
Wookje Han, Jinsol Park, Kyungjae Lee
EMNLP, 2023 [Paper]

Meta-Learning of Prompt Generation for Lightweight Prompt Engineering on Language-Model-as-a-Service.
Hyeonmin Ha, JIHYE LEE, Wookje Han, Byung-Gon Chun
Findings of EMNLP, 2023 [Paper]

A Survey on Memory Optimization Techniques and Frameworks for Training Large Language Models
Wookje Han, Taehyun Lee, Hyeonmin Ha, Byung-Gon Chun
KSC, 2022

Plug-and-Play Adaptation for Continuously-updated QA
Kyungjae Lee, Wookje Han, Seung-won Hwang, Hwaran Lee, Joonsuk Park, Sang-Woo Lee
Findings of ACL, 2022 [Paper]

---

Teaching

---

• Fall 2022 TA, Computer Architecture

      Seoul National University, Prof. Jin Soo Kim
      Best Undergraduate TA Award

• Fall 2022 Tutor, Computer Programming

      Seoul National University, Prof. Young Ki Lee

Last Updated: Aug 18 2026