My research focuses on the gap between technical evaluation and real-world consequence: who is selected, what knowledge is missing, how capacity changes outcomes, and whether a system remains trustworthy once it enters an institution.
The domains differ, but the research logic is consistent: aggregate performance is not enough. We need to understand allocation, completeness, context and the institutional system around the model.
Two systems can report similar aggregate performance while allocating scarce opportunities, risks or errors to different people. RIDI makes that hidden allocation structure visible.
Open RIDI interactive demo →Clinical and biomedical systems are only as trustworthy as the knowledge they can see. My work on drug–drug interactions, pharmacovigilance and personalized medicine examines how heterogeneous or incomplete knowledge changes what can safely be inferred.
See doctoral foundations →AI-assisted education needs more than generated content. The research challenge is whether learning design preserves evidence, progression, readiness and disciplinary fidelity when AI becomes part of the production system.
Explore teaching systems →Universities and public institutions face the same structural problem as AI systems: a metric is not a decision. My work examines operating models that connect evidence, merit, capacity, partnerships and accountability.
Discuss institutional research →The next stage is not another metric in isolation. It is a stronger evaluation architecture for systems that act under constraints.
I am open to joint studies, replication, evaluation and applied research partnerships.