Agent reasoning and evaluation · Post-training · Synthetic data
I’m a senior applied scientist at Microsoft working on AI agents for Copilot. I turn synthetic data and human feedback into evaluations and training signals that help agents become more useful and reliable.
I completed a Ph.D. in Human-Computer Interaction at Carnegie Mellon University. I’ve also spent time at MIT, the Allen Institute for Artificial Intelligence (AI2), and Conservation X Labs, where I helped develop team-formation tools for global environmental innovation contests.
I develop agentic quality graders, using controlled synthetic variations in output quality and calibrated comparisons to probe evaluation reliability and quality tradeoffs. I also evaluate tool-use effectiveness and efficiency, connecting agent behavior with task outcomes and computational cost.
I work on high-quality synthetic training data and reward models, supporting supervised fine-tuning and reinforcement learning to improve agent reasoning, tool use, and grounded responses.
I study how agents can resist manipulation and protect sensitive information. My work includes pipelines that generate diverse adversarial tests from everyday user requests to benchmark agent robustness, and HypoVeil for privacy-aware collaboration.
I build systems that help people decompose problems, discover ideas across domains, and synthesize scientific literature. Projects such as BioSpark, Synergi, and Threddy combine human framing and interpretation with AI retrieval, abstraction, and synthesis.
* Equal contribution
Earlier version: CMU Agent Workshop · 2025.
Best Paper Honorable Mention
Best Paper Award