To advance perspectives by bringing systems thinking and people to the centre of everything I do.
I work at the seam of research, design, and engineering. I author the research, design the experience, and ship the production code. Doing all three keeps the loop tight and the translation losses small, but it only matters if the work stays honest. These four principles are how I keep it that way.
Show the evidence and the reasoning, not just the conclusion. Findings should be auditable and defensible. Every claim traces to a source, and confidence is scored rather than asserted. When I build AI into a product, I expose how it reached its answer so a person can trust, check, and override it.
Ask before assuming. I treat research as infrastructure. Jobs-to-be-Done and outcome-driven methods, interviews, shadowing, and measurement. Decisions rest on what people actually need, not on opinion. The best questions reframe the problem before anyone touches a screen.
Translate chaos into manageable order. Good design eliminates the blank page: clear information architecture, a confident hierarchy, and interfaces that say what they do. If a person has to think about the tool instead of the task, the design isn’t finished.
Build the better version, then ship it. I treat AI as a design material. It's a powerful, probabilistic raw material to shape with guardrails and a human in the loop. The goal is experiences that feel magical and stay accountable: the AI proposes, the person decides.
These aren’t posters on a wall. Transparency is the audit layer in my AI research work at Articos. Inquiry is the PACT process behind Careem and the JTBD tooling at Strategyn. Clarity is the redesigned AutoLeap workflows. Optimism is shipping a human-in-the-loop AI repair-order flow to production. I write about the throughline in Writing.