01 Selected work
2009—2026
Enterprise systems,
clinical platforms,
design practice.
15+ years across healthcare, banking, enterprise software, and agency work. The constant: absorbing complexity from system constraints and keeping it invisible from the people who have to use them.
Current · Enterprise Healthcare
Where decisions
have consequences.
Clinical decision support, remote critical care, and analytics for high-stakes environments.
View case study ↗CareLogic: Clinical Decision Support Platform
A three-application platform for clinical decision support used by healthcare organisations globally. Designed the protocol lifecycle, configuration systems, and embedded analytics, entirely through architecture and process, no screens under NDA.
View case study ↗Mural: A Canvas for Tele-Critical Care
A remote-ICU platform where one clinician monitors whole ICUs from a central station. The whole interaction model follows from a single distinction: remote, not bedside. Observe and escalate, never act on the device.
View case study ↗Clinical & Operational Insights
Tactical and strategic dashboards for ICU and Labor & Delivery, built as a mirror rather than an engine. The project I learned the most from, because it stalled, and what I'd do differently in its second life.
Design leadership · The Minimalist
The product was
the practice.
Four years designing the mandate, method, talent system and commercial model behind a multidisciplinary design organisation.
Human factors · HFI
When the human fails,
look at the system.
The foundation: mental models, language, constraints and the discipline of refusing “user error.”
View case study ↗Standard Bank IVR: South Africa
Redesigned a telephone banking IVR for semi-literate users in South Africa. Audio-only, no visual affordances, zero tolerance for errors. A proto-agent design problem solved with card sorting, mental model mapping, and Wizard of Oz testing.
View case study ↗HDFC Internet Banking Simplification
Five different labels for the same task because the data lived on different systems. The constraint: we couldn't change the backend. The solution: absorb the routing at system level and give the user one label, one interaction.
After hours