Untangling complex products is my favourite job
For 10+ years I've designed the products most people find intimidating: AI marketing platforms, football scouting systems, airline design systems.
My job is the moment intimidation becomes obvious: when a marketer reads a media-mix model without an analyst, or a scout files a report from a stadium seat without thinking about the interface at all.
Research that changes decisions
Enterprise platforms, analytics tools, multi-user systems. I think in flows, states and information architecture before pixels, and I'm happiest untangling the products everyone else avoids.
Leadership
I built a UX department, I mentored designers across a 350-person programme, I lead by raising the bar of the work itself, not by leaving the tools behind.
AI UX, done responsibly
I've designed AI summaries, explainability patterns, and human-approved recommendation flows for enterprise AI. My rule: AI earns its place in a product by being useful and accountable, not by being the headline.
How I think
1. Readability first. Then attention. Then intelligence.
If people can't read it, nothing else matters. Most teams build this pyramid upside down: leading with AI, then wondering why nobody trusts the product
2. Adoption is the only metric that survives contact with users.
A tool that ships but isn't used is a beautifully documented failure. I design for the sceptical expert: the scout, the marketer, the analyst who already has a way of working and didn't ask for a new one
3. Evidence beats opinion.
Including mine. Interviews, audits and testing exist to make decisions cheaper than guessing
4. Systems, not screens.
Any single screen can be made beautiful. The craft is in the hundred screens you haven't designed yet
5. AI should behave like a good colleague.
Useful, transparent about how it reached a conclusion, and comfortable being overruled. When I design AI recommendations as accept / customise / ignore, that's not a UI detail — it's a statement about who's in charge.
6. Restraint is a skill I learned from art.
Fine art taught me to recognise when a piece is finished, which is earlier than most teams think. Knowing what to leave out is the difference between an interface and a control panel.
Research & Discovery Approach
Good research is cheaper than confident guessing. Across a decade of complex products, I've used research to decide what to build, settle stakeholder debates with evidence, and catch expensive mistakes while they were still sketches.
Methods, plainly:
Stakeholder interviews · User interviews · Discovery workshops · Heuristic evaluation · Design & accessibility audits · Usability testing · Competitor analysis · Journey & service mapping · Synthesis & insight reporting · Research ops for distributed teams
Stakeholder interviews, contextual user interviews, discovery workshops, service mapping. Used at: City Football Group (understanding how scouts and executives actually plan squads), Unilever AI MMM (workshops with market teams to define what decisions the data must support).
When we have a product but not confidence → heuristic evaluation, design audits, accessibility audits, analytics review. Used at: Unilever Accessibility Hub (auditing experiences across a ~400-brand organisation), Healthspan, Thinkbox.
Usability testing, prototype walkthroughs with real users, iterative feedback rounds with market teams. Used at: SheaMoisture & Degree discovery phases, CFG (validating the planner with the people who run transfer windows).
Research as referee: synthesis sessions where findings, not seniority, settle the argument. Used at: everywhere, honestly. This is half the job.
I also make art
I'm a practising mixed-media fine artist. My work has appeared on the 10th-anniversary cover of Tempus magazine, and I've spoken on their panel about art and craft.
This isn't a hobby footnote. It's why my systems work doesn't look like systems work; a trained eye for composition, colour and restraint applied to enterprise software. The intersection is the point.
