I study why AI systems fail the people who depend on them — and how to design them so they don't.
With a background spanning hospital IT, clinical data analysis, and academic research, I work at the intersection of healthcare informatics, human factors, and responsible AI. My focus is on the gap between how AI is built and how it is actually used — in clinical workflows, organisational processes, and real-world deployments.
I spent over seven years managing servers, web infrastructure, and IT systems, including inside a clinical environment at Mashhad University of Medical Sciences. I was close enough to healthcare to see how data moved — or failed to. Systems were designed without asking the people who used them. Workarounds were everywhere. Data sat in silos. Reports reflected what was easy to record, not what was clinically meaningful.
Why do technically sound systems get abandoned, gamed, or ignored in healthcare? I moved into Medical Informatics and then Computer Science research to find a structured answer. At UCD, working on ontology engineering, socio-technical systems analysis, and human factors, I found that the failure point is almost always the same: the system was designed around data flows, not around people and their context.
During my placement at St. James's Hospital Dublin, I analysed data collection, quality management, and reporting practices across a major acute hospital. I documented data quality gaps, mapped fragmented workflows, and produced structured outputs for clinical and senior operational stakeholders. The problem wasn't the data — it was the mismatch between how the system was designed and how clinical work actually happens.
Across the VICTORY and ID4AI projects at UCD, I contribute to healthcare AI innovation, inclusive design research, and AI governance. My work sits at the intersection of three things: understanding how AI systems interact with real clinical workflows, ensuring data is transparent and trustworthy, and helping organisations think through what responsible AI deployment actually requires — not just technically, but organisationally and humanly.
Not an AI developer. Not a pure data scientist. Something more specific.
I map clinical and organisational systems to identify where data quality breaks down, where accountability is unclear, and where AI interventions will create friction rather than value.
Using socio-technical systems analysis and human factors methods, I assess how AI systems build or erode trust inside organisations, and what governance structures support responsible adoption.
I work across multidisciplinary teams — healthcare, research, and technology — translating between what clinicians and users need and what systems can realistically deliver.
A scoping review examining how AI systems in healthcare build or undermine organisational trust — using socio-technical systems analysis to go beyond technical compliance.
Mapped healthcare data lifecycles using PROV-O and SEIPS to make data provenance visible and auditable — addressing the gap between what systems record and what clinicians actually need.
The intersecting fields that shape my work.