About me

I’m an Applied Scientist at Amazon working on large language models, retrieval, and agentic systems for real-world applications. My research focuses on how intelligent systems acquire, preserve, compose, and seek knowledge. I am interested in understanding how knowledge can be learned efficiently, retained without forgetting, connected across multiple sources, and obtained when it is not immediately available. More broadly, I study how these capabilities support reasoning, planning, and decision making in intelligent systems.

My work spans both foundation models and production AI systems. On the Prime Video Search team, I develop language-model-based systems for real-time query understanding and agentic search under strict latency and scalability constraints. This work explores how models can combine internal knowledge with external information sources to better understand user intent and solve complex information-seeking tasks.

Previously, on Amazon’s Stores Foundational AI team, I worked on domain-specialized language models for shopping applications. My research focused on continual pretraining and domain adaptation, studying how new knowledge can be incorporated into foundation models while preserving their existing capabilities. These models were adopted across multiple product teams and supported a wide range of customer-facing applications.

Before joining Amazon, I received my Ph.D. in Computer Science from the Korea Advanced Institute of Science and Technology (KAIST), advised by Alice Oh. My doctoral research focused on question answering, retrieval, and representation learning, with an emphasis on connecting information distributed across multiple sources. Across both academia and industry, a common theme of my work has been understanding how intelligent systems acquire, organize, and utilize knowledge to solve increasingly complex tasks.