I design AI products people can check, trust and ship. Most recently at Liminal, where I took two review tools for AI agents from whiteboard to production.
Before that: websites at Digital Silk, an AI sales platform and its design system at Aptivio, and client products at Wunderman Thompson.
→ Hello
→ Experience
→ What I bring…
Product designer for AI and B2B SaaS. This is how I help teams ship the right thing, sooner.
- AI product design
AI that people can trust
I design the review steps that keep humans in charge of AI output. Two shipped at Liminal, each in about five weeks.
- Prototyping in code
Specs developers can run
Working React prototypes on the real design system, so less gets lost between design and build.
- Design systems
Consistency that scales
Tokens, components and patterns that let teams ship faster without drifting. I led the Boon design system.
- UX research
The right problem first
Workshops, interviews and data before screens. One workshop turned a menu redesign into an audit trail and a QA step.
- Iteration
Feedback into decisions, fast
After one launch I turned 23 pieces of user feedback into 15 design decisions in five days.
- Collaboration
Smooth work with engineering
Ticket-ready specs, short videos and numbered questions keep teams moving, even across time zones.
- Quality and accessibility
Quality checked, not assumed
I check what got built against what was agreed, WCAG included. One audit found 24 agreed changes with no ticket.
- AI-native workflow
Faster from idea to users
I use Claude Code and agents for prototypes and audits. One change to an AI connector cut tool calls by 60–80%.