My research centers on long-horizon human–AI interaction. I design backbone frameworks and models for agentic systems that ingest streaming, multimodal inputs; manage and retrieve persistent memory; and plan and reflect over extended tasks — making AI collaborators reliable, adaptive, and self-improving.

I am currently an Applied Scientist at AWS (Agentic AI), where I lead work on intelligent LLM routing for Amazon Bedrock and vision language action models.

Previously, I earned my Ph.D. in Electrical Engineering from Yale University, advised by Prof. Leandros Tassiulas, where my research focused on networking and distributed learning. I completed a B.Eng. in Optoelectronics at Zhejiang University (ranked 2/124) in the Advanced Honor Class of Engineering (ACEE) at the Chu Kochen Honors College, and spent a year on exchange at HKU with research internships at CUHK and UIUC. I have also interned at Microsoft, IBM, and Nokia Bell Labs.