Product designer — systems, services, agents
Deep roots in design systems, facilitation, and research, applied across enterprise ecosystems, hyper-growth startups, and solo products. These days that practice extends to agentic workflows and the sovereign infrastructure they run on.
Service design, data visualization, design systems, and full UX — built from scratch and in active use. 55 reports for 20 clients so far; production time per report down from ~8 hours to ~3.
Responsive web, account redesign, omnichannel design system, COVID-era kiosk library — 80+ shared components across 6+ surfaces, and 134 kiosk screens rebuilt in a four-week solo sprint.
A research investigation into how language, systems, and assumptions shape the experience of disabled air travelers. Cross-method triangulation, advocacy-network recruitment, and the single-word insight that drove a service rebrand and training redesign.
No dedicated PM, high turnover, headcount doubling — and a 56-component library that every product squad adopted in its first year, used by 112 designers and engineers.
Six sessions, six participants, one living playbook the team still owns. The “interface” was a room. The “product” was a changed way of working.
Expressive storytelling, data visualization, cross-agency collaboration. Shipped during the first pandemic holiday season.
Governed ~50,000 assets and 500 registered users on the enterprise DAM for Intuit's Mailchimp marketing org, expanding into QuickBooks and Credit Karma workstreams.
I start by making the system legible: who is involved, what they are trying to do, where risk or friction concentrates, and what decisions the interface needs to support. Research gives the shape of the problem; prototypes and system models test the decisions before they become production commitments.
Map actors, workflows, constraints, edge cases, and success criteria before narrowing into screens.
Use interviews, usability studies, artifact audits, journey maps, and domain research to separate signal from assumption.
Turn tradeoffs into artifacts: flows, prototypes, principles, taxonomies, service blueprints, and reusable patterns.
Document the why, contribute to the system, and create governance so teams can apply the decision without me in the room.
I design and run multi-agent pipelines for real work — research synthesis, ops automation, and an end-to-end job-search system with structured data contracts. That practice extends into AI-native design tools: Paper, Pencil, Agentation, and Figma over MCP. The governance thinking I bring to design systems applies to agents too: clear contracts, documented decisions, systems that work without me in the room.
claude code · figma mcp · paper · pencil · agentationThe era of abundant software is all about the homemade — building intimate, sovereign digital spaces rather than renting them. I run the platforms my work lives on: self-hosted publishing, data migrations, automation on personal infrastructure. I want to build software that feels like entering a friend’s living room.
ghost · cloudflare · homelab · sovereigntyBuilt agent-first: semantic HTML, structured data, and an llms.txt so both crawlers and AI agents can read it without a headless browser. Designed and shipped in a human–agent pair, the same way the rest of my practice runs.
next.js · json-ld · llms.txt