Work

Self-authored clinical risk assessment applying the NHS DCB0160 framework to an AI-integrated clinical workflow. Written independently to understand what responsible AI deployment in a clinical setting actually requires in practice — not to produce a certified document, but to do the thinking the framework demands. The process surfaced assumptions about AI risk that the clinical safety literature handles very differently from the mainstream AI safety discourse.

MSc Psychology dissertation examining the psychological predictors of engagement with AI-driven health coaching tools: trust, perceived usefulness, privacy concern, and behavioural intention to use. Small sample study using structural equation modelling. The finding that has stayed with me: adoption failure in AI health tools is rarely about the technology and almost always about the relationship between the user's mental model and what the tool actually does.

Authored a practical guide mapping 19 real AI use cases for community dietitians working within NHS Wales. Not theoretical — each use case was grounded in the actual tasks that consume time in a clinical day: letters, summaries, patient-facing materials, administrative burden. The goal was to bridge the gap between what Microsoft Copilot can do in principle and what a community dietitian actually does between 9 and 5.

Designed and built a patient management platform for a community dietetics service using Microsoft Power Apps. Built around actual clinical workflow — not an assumed one — with input from the team who would use it. A formative lesson in how much the distance between "what we designed" and "how it gets used" depends on whether you involved the right people before you started building.

Evidence-based usability review of a public community space for neurodiverse users, applying accessibility and user research methodology to a real-world physical environment. Proposed concrete, costed design modifications grounded in the literature on sensory processing and spatial cognition in ADHD. An early exercise in translating research evidence into practical design recommendations.

Designed and delivered a structured behaviour change course for patients exploring the relationship between emotion and eating. Applied COM-B model and motivational interviewing principles to a digital video format. Built from a clinical observation I kept returning to: the patients who make lasting change are rarely the most informed — they're the most clear about why change matters to them specifically.