Yingqi Huan is a master’s student in Applied Statistics at Teachers College, Columbia University. Her research investigates how generative AI can be leveraged to support diagnostic feedback in second language (L2) writing assessment. Rather than focusing solely on automated scoring, she aims to develop and evaluate models that provide interpretable, formative feedback, helping learners understand why their writing performs well or poorly across diverse dimensions. Her broader research integrates causal inference and psychometric modeling to identify how feedback affects learning trajectories and to design equitable, causally grounded assessment systems that bridge the gap between measurement theory and AI applications.