Selected work · 2025–present
Case studies in product strategy, complex workflows, and AI-assisted product development.
Each entry generalizes proprietary work into a portfolio-safe narrative. The artifacts shown are conceptual; the thinking, decisions, and outcomes are real. All of the UI shown here is shipped and in active use by engineering teams at some of the world's largest enterprises.
- 01→
MCP Explorer
Designing a surface where humans and agents share the same vocabulary.
A discovery and reasoning interface for the Model Context Protocol, built so engineering teams can make their tools legible to AI agents safely.
- 02→
Governing Design Drift
Three plugins that put one source of truth in front of designers, in whatever surface they already work in.
Three plugins built on one idea, that code is the source of truth and the design tool is a projection of it: one reports how the app uses the system, one reconciles a design file against the spec, and one publishes it into a command-line design system.
- 03→
A Theme Explorer for Docs and Platform
A sandbox for design, brand, and engineering to try color, type, and light or dark live, and align on a direction in minutes instead of weeks of hand-built mockups.
Exploring a new visual direction for our Docs and platform used to mean rebuilding color and type options across many screens in Figma, weeks per round. I built a theme explorer on our team's UI kit and the marketing site's architecture, so design, brand, and engineering can swap color, type, and light or dark live, watch every screen re-theme, and align on a direction in minutes.
- 04→
AI-Oriented Recipe Documentation
An information architecture that serves humans, AI systems, developers, and product goals at once.
Rethinking recipe pages as a shared interface layer between developers, AI agents, and the product, using only the data that already exists.
- 05→
Designing a Changelog for Constant Change
A single, organization-wide view of every commit and pull request across all repositories and source control systems, with status and bulk actions in one place.
A brand-new, organization-wide view that aggregates commits and pull requests across every repository, designed from a mental model and Figma, then hardened with AI so the UI held up when real data arrived.
- 06→
Space Cop: A Space Situational-Awareness Dashboard
A command dashboard for orbit analysts tracking launches, re-entries, breakups, and on-orbit objects. Designed at Tanzu Labs (formerly Pivotal Labs) with weekly user research.
Space Cop gives space-operations analysts one risk-first view of launches, re-entries, breakups, and on-orbit objects, with a tracking map and a responsive detail view. Designed during a Tanzu Labs (formerly Pivotal Labs) engagement, it was shaped by weekly user interviews and a balanced-team feedback loop.
- 07→
Artifact Deployment Workflows
Designing for how users think about deployments, not how they map to a table.
Reorganizing a deployment surface around the concepts users actually work with: ecosystem, version selector, deployment state, and action.
- 08→
Docs Homepage and a Constrained IA Redesign
An information architecture and homepage redesign that cut a long, deeply nested nav into clear entry points, with every page left at its original URL.
A manual IA and homepage redesign for a large developer-docs site, constrained to preserve every URL. Done by hand, and an honest account of where AI would have made the audit faster.
- 09→
Restructuring Navigation for a Wealth Advisory Platform
A cross-team workshop set the strategy. I led the execution, grounding a scalable navigation model in how advisors actually think about clients and accounts.
New integrations were straining a navigation model that mixed places, actions, and workflows in one rail. A workshop with design leadership and research set the direction, then I led the redesign into a structure that nests accounts under clients and moves contextual actions into a dedicated panel.
- 11→
Roadmapping by Complexity, with AI
A tactical way to use AI for fast initial planning: inventory what exists, sort every idea by what it costs to build, and the MVP becomes obvious.
A designer's tactic for going from a long wishlist to a defensible MVP roadmap in an afternoon, using AI to inventory reusable parts and sort each idea by complexity. A starting point to validate with the team before it becomes a real plan.
- 12→
Building a Portfolio with an AI Agent
A living system: capture private work, transform it with a constrained agent, publish under a safety charter.
How I built this portfolio as a governed human-AI system, and a playbook for designers who do confidential work and still need to show it.