Hey,
I'm Karlee
I'm a Principal Product Designer specializing in design systems — building token architectures, component libraries, and the infrastructure that lets design and engineering move as one. Currently on the Fluent Design System team at Microsoft, leading Fluent Flex Tokens across M365.
I grow systems, from the people in them to the people they reach.
There's a System Under the System (I Built It)
Built so every downstream artifact is an output of the system, not a parallel effort. Design intent lives in the token name — meaning the system scales to platforms, themes, and AI without touching component logic.
I've Never Designed Anything Like This
A markdown-first repo where one spec file simultaneously serves as designer reference, engineering source, and AI agent prompt — with headless component delivery and a 7-minute Figma build via Claude.
From Fork to Framework
Delivered Copilot as a token-based theme extension of Fluent — not a parallel system. Full code-to-design parity across 72 assets and a brand theming framework now used by 10+ Microsoft products.
Before Anyone Touches a Component
Before the work starts, I build the infrastructure for how it will get done. Audit frameworks, measurement systems, a process that holds across a distributed team — turning ambiguous scope into a shipped plan.
Where It All Started
The project that first made me think in semantic layers. A brand color update request revealed a brittle system — so we rebuilt it with a semantic variable layer that turned hundreds of future updates into a handful.
Component + Property Naming Framework
Naming things is hard. A research-backed framework — two workshops, 11 categories, card sorting and pressure testing — standardized how Homebase names components and props across design and code.
Overlays and UI Regions, Oh My!
A community proposal to update the Drawer component turned into an audit of 17 design systems, a hand-calculated box-and-whiskers analysis, and the conclusion that what teams actually needed was a brand new component entirely.
"She sequences work so that we're prepared and not blindsided down the line. Her work was instrumental in helping us build a system that has longevity and flexibility — and it's going to be foundational for all of our AI efforts moving forward."