I am an M.S. student at the IRIS Lab, Sungkyunkwan University (SKKU), advised by Prof. Jong Hwan Ko. My research focuses on efficient AI inference, including diffusion large language models (dLLMs), KV-cache compression and quantization, hybrid attention, and hardware-software co-optimization. I am particularly interested in making large generative models faster and more memory-efficient through algorithm and system-level design.