Profile photo placeholder for Kangyoon Lee
Replace with your own portrait before launch

Kangyoon Lee

Ph.D. Student in Computer Science

Johns Hopkins University · Baltimore, MD

Research Interests

Distributed AI Communication · AI Datacenter Networking
GPU Collectives · Workload-Aware Routing & Load Balancing

I am a first-year Ph.D. student in Computer Science at Johns Hopkins University. My long-term vision is to enable everyone to independently choose, customize, and deploy AI. To work toward this goal, I am interested in understanding AI workloads and optimizing the systems and resources that support them.

Specifically, I study communication in distributed AI systems, including the origins of traffic bursts during collective operations, with the goal of informing more effective scheduling and load balancing. Before joining JHU, I was an AI Research Intern at the Korea Institute of Science and Technology (KIST). I received my B.Eng. in Information and Communications Engineering from Hankuk University of Foreign Studies in 2025.

Selected Projects

All projects ↗

Recent News

All news ↗

Selected Publications

All publications ↗
2026 · IPIU · Poster
Frequency-Aware Multi-Resolution Architecture for High-Resolution Spatiotemporal Map Reconstruction from Sparse Observations
Eunshik Kim, Kangyoon Lee, Yekwon Kim, Hyungjoo Jung, Heeseung Choi, Igjae Kim, and Haesol Park
2025 · IEIE · Poster
Learnable Adjacency Matrix Based Spatio-Temporal Graph Convolutional Network for Traffic Prediction
Yekwon Kim, Eunshik Kim, Kangyoon Lee, Haesol Park, Heeseung Choi, Igjae Kim, and Hyungjoo Jung
Paper link: [TBD]