I joined the Reality Labs design systems team during a critical build phase, working alongside the production design lead on an at-risk work stream for RLDS 1.1. My contributions spanned the creation and maintenance of the design system infrastructure within Figma — covering tokenization, component work, templates, and documentation — as well as broader initiatives around designer education, design critiques, and AI tooling to accelerate prototyping.

• Design system maintenance in Figma across tokens, components, and templates
• Established Code Connect connections between Figma components and production code
• Worked on a Design Linter to enforce system consistency at scale
• Built a prototyping library enabling designers to build across the 4 engineering frameworks supported by our partners
• Contributed to designer education and design critique processes

The Reality Labs Design System is the token, component, and documentation foundation for Meta's VR platform. VR is a fundamentally different design problem than mobile. The platform spans mixed-reality surfaces with few established conventions, multiple product features across a growing hardware lineup, and tight, hardware-bound release deadlines that leave little room for design debt.

When I joined, the team was at a pivotal moment — scaling the system to support dozens of designers and engineers across Reality Labs while keeping the library coherent and maintainable by a small core team. The challenge wasn't just building components. It was building infrastructure that wouldn't collapse under the weight of the platform it was meant to support.

The work centered on three interlocking problems that needed to be solved together rather than in sequence.

First, the library needed to scale across mixed-reality surfaces — a surface area with no settled playbook and wildly different interaction paradigms from screen to screen.

Second, the token architecture needed to carry theming and density variations without forking components. Forking is a design system antipattern that compounds over time; the goal was a single component tree that could adapt through tokens alone.

Third, the system needed to be approachable enough for dozens of designers and engineers to build on it without every question escalating to the core maintainers. This shaped how we approached documentation, Code Connect, the Design Linter, and the prototyping library — all of which were about reducing the surface area of confusion for contributors.

Product designers across Reality Labs used RLDS as the shared foundation for building and shipping VR product features. With Code Connect established, designers could move from Figma components to production-accurate prototypes without needing to negotiate the gap between design and engineering by hand — a particularly meaningful unlock when working against hardware release schedules.

The prototyping library gave designers a way to build across all four of the engineering frameworks our partners supported, removing the guesswork about which patterns would translate and which wouldn't. The Design Linter enforced system consistency automatically, which meant the small maintenance team could focus on building new infrastructure rather than auditing existing work.

Design systems work at this scale is hard to attribute to a single metric, but the goal was clear: give a small team the leverage to support a large one without the system becoming a bottleneck. The combination of Code Connect, the Design Linter, and the prototyping library moved the team in that direction — reducing the coordination overhead between design and engineering and making it easier for product designers to work with confidence that what they were building would hold up in production.

Working on RLDS sharpened my understanding of what design systems actually need to do at scale. The VR context made the stakes more visible — there's less precedent to fall back on, the surface area is genuinely novel, and the release cadence is unforgiving. That combination forces rigor around decisions that might otherwise get deferred.

It also reinforced something I've come to believe more strongly over time: the most important output of a design systems team isn't the component library — it's the infrastructure that lets other teams use it well. The Linter, Code Connect, and the prototyping library were all expressions of that idea.

Screenshots and GIFs showcasing components, specs, and templates from my work on the Reality Labs Design System.