Systems that build the thing.
A design system is a field with rules — the same forces, organized into tokens, components, and behavior a team can run. These are the reusable systems underneath the work.
Product systems
AI systems
Applied AI and AI interfaces: LLM tooling, computer vision, semantic search, and Model Context Protocol servers — built so the model output can be trusted, not just generated.
Commerce systems
Marketplace and storefront architecture: seller verification, product customization, payments with Stripe Connect, and AI-assisted discovery on one shared commerce core.
Interaction & motion systems
Interface behavior treated as a system — state, feedback, transitions, and rhythm captured as reusable motion primitives that hold from brand guidelines to code.
Manufacturing systems
Physical production as an operating system: 3D modeling, engineering drawings, in-house manufacturing, and fulfillment. A rare pairing with the software side.
Trust & verification systems
Typed contracts, schema validation, and the transformation layer that turns unreliable output into artifacts a downstream system can trust. The load-bearing idea behind much of the AI work.