Seven years in product. This is what good actually looks like.
Seven years designing inside IBM product teams — from a ground-up redesign of IBM MQ at the Carbon 9 to 10 transition, through five products and counting, to leading design on Asset Explorer today. The through-line is craft under real delivery pressure: beautiful, detailed experiences built fast enough to matter.
Exemplars only work if they map to the real jobs product teams face. I'm not approaching Carbon Next from the outside — I'm arriving with seven years of firsthand experience of what designers and developers actually reach for, where systems let them down, and what it takes to ship quality at pace.
The story
Asset Explorer GA
IBM webMethods Hybrid Integration spans 11 capabilities, each producing assets with value beyond their original context. But there was no unified way to discover, understand, or reuse them — valuable work stayed hidden behind capability silos and tribal knowledge.
As the sole designer, I led Asset Explorer from concept to GA — a cross-platform discovery experience for the ecosystem. I defined a realistic MVP from an ambitious vision, aligned stakeholders, made difficult prioritisation calls, and advocated for the experiences that would deliver the most value first.
AI fluency through process
Joining a newly acquired team as the sole designer, I faced a familiar scaling problem: too many visual defects, too little time before release. Manual review and Jira workflows weren't keeping pace.
I introduced AI-assisted fixes directly within the codebase — fitting into established engineering workflows rather than disrupting them. I became the first designer in the organisation to contribute production code this way, then shared the approach across the wider design org. The result: a faster path to quality and a stronger design–engineering partnership.
AI fluency through experience
I'm now exploring how IBM webMethods Hybrid Integration can operate inside the AI tools users already rely on — giving agents the context they need to act more like knowledgeable employees.
My current focus is a context engine: enabling asset discovery and reuse through a headless experience, while shaping UI patterns for building context and endorsing AI-generated enhancements. It's also surfaced a clear need for something like Carbon Next — our current design system isn't built for experiences shared between human input and machine action.