I design AI systems that act on behalf of users — diagnosing problems, proposing fixes, and executing within defined boundaries. The design challenge is making that feel trustworthy and controllable, not opaque. I solve that by building transparency and reversibility into every autonomous action.
18 years across security platforms, fintech, and enterprise SaaS. Currently at Elastic, where I'm designing agentic detection workflows — systems that diagnose broken rules, propose fixes, and execute autonomously within analyst-defined scope.
Before that: founding designer at HouseToken (blockchain real estate), lead designer at Keysight Technologies (AI test automation), and five years at Stanley Black & Decker designing across European markets.
I work best on products where the domain is hard and the users are expert. The less forgiving the user, the more precise the design has to be.
Experience
Designing agentic detection workflows for Elastic Security — systems that monitor rule health, diagnose failures, propose fixes, and execute autonomously within analyst-defined autonomy settings. The product (AutoDEX) achieved a 91% analyst approval rate and 34 autonomous actions in the first week of beta across three enterprise accounts. The design challenge was trust architecture: how do you make a system that acts faster than users can verify feel controllable rather than unpredictable?
Led end-to-end redesign of the Eggplant AI Modeler — migrating from a legacy Dojo framework into a modern React SaaS platform. The core problem: users were mentally translating system terminology into their own application model, opening multiple windows to cross-reference what they were building. The fix: embedded real application screenshots directly into the modelling surface. Outcome: 28% reduction in model creation time for first-time users, measured in task-based usability testing with 12 participants. Also managed the global design system and ran the UX research programme across qualitative and quantitative methods.
Sole designer, zero-to-one. Designed the mobile investment platform from brand identity through to shipping — research, information architecture, interaction design, and engineering handoff. The hardest problem: users wanted to invest in property, not cryptocurrency. The solution was to abstract blockchain entirely and lead with return visibility. 90% of first investments were completed after users saw projected returns in the first screen of the listing.
Senior designer across European markets, leading digital initiatives for multiple brands including Black & Decker. Ran cross-market user research across focus groups in five European countries — UK, Germany, France, Netherlands, and Sweden. Presented design strategy to senior stakeholders across the organisation. Led the UX research and redesign of the connected robotic mower app, restructured around action-first hierarchy based on research findings.