Research Portfolio
Two Threads, One Question.
How do we build AI systems that truly understand human intent? I approach this question from two complementary angles: formalizing the evaluations AI must master, and designing the interaction paradigms that keep humans in control.
From Fluency to Formalization
AI Alignment & Semantic Grounding
Building structured representations—taxonomies, benchmarks, and intermediate languages—that force AI systems to reason about the effects of their actions, the quality of their outputs, and the context of user intent.
Beyond Visual Defaults
AI for Accessibility & Human Agency
Building AI systems that decouple creativity from sight—transforming accessibility from a compliance checkbox into a blueprint for Multimodal, Verifiable, and Customizable Human-AI collaboration.