Skip to main content

Work

One throughline, four moves. A progression across Amazon Fulfillment Technologies (AFT): from defining shared UX foundations, to unifying the knowledge behind them, to delivering that intelligence directly into builders' workflows, to closing the loop with real usage.

01

Define the standard

Built the shared foundation teams reason from: design-system and token architecture, a framework-agnostic UX standard across a 500+ product ecosystem, and the tooling that made it self-serve. Built for its moment. The org has since moved to Amazon's CCR/WCAG accessibility baseline, and the real leverage moved to scaling how teams apply standards at all, which is the work of stages 02–04.

02

Unify the knowledge

Consolidated fragmented research, standards, and fulfillment operational context into one queryable source of truth, so any builder could answer “what do we already know?” consistently.

03

Distribute intelligence into the workflow

Flagship

Purpose-built AI agents, delivery through the Model Context Protocol, and a repeatable design pipeline that puts shared guidance directly inside the tools builders already use, instead of asking them to visit another destination.

04

Close the loop

In progress

Feeding real product-usage signal back into the system so guidance reflects how tools are actually used. The analytics integration is live in production, and the guided plays that turn that data into decisions are rolling out now.