I build AI agent systems.

I’m a Senior Staff Researcher at Alibaba, based in Bellevue, Washington. Our team works on two fronts: probing the limits of what LLMs can do through new applications and agent harness design, and making them production-ready through training platforms and secure runtimes. Current projects include QwenPaw, a self-improving and secure personal assistant; DojoZero, a platform for AI agents that reason and act on real-time data, such as predicting sports outcomes; and TuFT, a multi-tenant platform for fine-tuning LLMs through a unified API. Our work is open-source under AgentScope.

Previously, I was a Principal Researcher at Microsoft Research in Redmond, Washington, where I was the lead architect of AutoGen , an open-source framework for building AI agents and multi-agent applications. AutoGen offers both a high-level API for orchestrating agents that collaborate autonomously or with human supervision, and an event-driven, low-level API that gives developers full control of agent behavior. The framework is built on the Actor Model, with a distributed runtime that can host agents written in different programming languages. I was fortunate to lead a vibrant community of 500+ contributors who shaped AutoGen into a widely adopted tool for developers and researchers inside and outside Microsoft.

I also joined forces with Azure AI to build the Foundry Agent Platform, a cloud platform for deploying AI agents and workflows at scale. As part of this effort, I helped create the Microsoft Agent Framework , an evolution of AutoGen for enterprise-grade applications and a core component of the Foundry Agent Platform.

Prior to working on AI agents, I worked on database systems at Microsoft Research. We developed a cost-based, platform-independent rewrite rule for MATCH_RECOGNIZE queries in general-purpose SQL engines, achieving a 5.4X median latency improvement in Trino. We then built a specialized execution engine for MATCH_RECOGNIZE with extended operators and a novel cost-model-based optimizer, delivering a 6X median performance gain over state-of-the-art engines.

Before joining Microsoft Research, I completed my PhD in Computer Science at the University of Toronto, advised by Prof. Renée J. Miller. My thesis focused on dataset search over massive Open Data archives. I contributed algorithms for large-scale set similarity search and data sketches , capable of finding joinable or unionable tables from over 100K tables in milliseconds. I also built an Open Data search engine stack to make Open Data more accessible for downstream applications.

I explore new ideas in computing through open-source projects and writing. My goal is to make advanced AI and algorithms accessible to everyone.

[Github] [Google Scholar] [Blog] [Email]