I am a Senior Applied Scientist at AWS Agentic AI, working on large language models (LLMs) and agentic AI. I build systems that enable language models to acquire and internalize knowledge, reason faithfully, and act as autonomous agents.

Previously, I was a Postdoctoral researcher at MIT CSAIL with Regina Barzilay, focusing on AI for chemical sciences. I received my Ph.D. from HIT SCIR, advised by Haifeng Wang and Wanxiang Che, and was a visiting student at Johns Hopkins University with David Yarowsky.

My recent work includes:

• Post-Training — Amazon Titan LLMs
• Text-to-SQL — Amazon Q in QuickSight
• IR / RAG — Amazon Bedrock Knowledge Bases
• Agent Memory — Amazon Bedrock AgentCore