I trained as a doctor and worked across humanitarian health programmes, where I saw how limited digital infrastructure made scarce resources harder to coordinate. Records were often incomplete, medicine stock and patient data were disconnected, and basic decisions relied on manual workarounds. That experience led me to begin learning how software and data might help. The projects below are how I am beginning to explore that interest, through health information, medical education and data analysis.
Health-Claim Verification Engine
A retrieval-augmented engine that splits sources across five models, surfaces their dissent, and expresses confidence as a range — not a false-precise number.
Virtual Patient
A simulated history station: question an in-character AI patient who reveals the case only when asked, then get a rubric-based examiner debrief on what you covered and missed.
Cerebral Perfusion & Cognition
A reproducible analysis of current glycaemic control, vascular brain injury and cognition in older adults with type 2 diabetes — pre-specified, honest about a small cohort.