I am an incoming M.S. student in Computer Science at Stanford University and a research trainee at Mila - Quebec AI Institute. My work focuses on reliable machine learning systems for high-stakes and socially consequential settings.

Current focus

• Multimodal medical AI: privacy-preserving MLLM agents for personalized multiple-sclerosis lesion-activity prognosis from clinical narratives and MRI.
• Trustworthy and interpretable learning: evidence-grounded workflows, model failure analysis, and multimodal decision representations.
• Computational social science: political-bias detection, misinformation analysis, and multi-agent LLM systems.
• Fair decision systems: resource allocation under diversity constraints and explainable graph learning.

At Mila, I work with Professor Tal Arbel on a local retrieval-synthesis-critique system and methods for identifying systematic failure subgroups. My previous research spans fair resource allocation and psychosis prediction at UBC, political-bias detection at McGill, and mobile glucose sensing at SFU.

| |