6 ways teams are experimenting with code layers
Camille Rose Shortridge · 23 October 2026
Read the originalFigma reports that product managers and designers at Cox Automotive, Dell and four other teams use coded prototypes on its canvas to align stakeholders before engineers build.
Camille Rose Shortridge, a customer marketing manager at Figma, the design tool maker, describes six teams using code layers. Code layers are prototypes built from working code on the Figma canvas with help from the Figma agent. The teams, which include Cox Automotive, Dell, Rocket Money, Merkle, Checkout.com and the creative agency Superside, mostly used the prototypes to win agreement before engineering work began.
- Lindsay Brechler, a senior product manager at Cox Automotive, produced 19 versions of a dealer feature with about an hour of her own time, the AI agent doing the editing.
- Miguel Wermuth, a product designer at Dell, builds prototypes from the same React components the developers use, so reviewers see how the shipped product will behave.
- Derek Onay, a lead product designer at Rocket Money, a personal finance app, used a coded prototype to settle interaction details and accessibility before the concept was approved.
- Merkle, a global consultancy, reports 30% faster iteration and up to 50% less time writing documentation such as interaction specifications.
Product managers and designers can show stakeholders a working prototype instead of describing behaviour in documents. This is Figma's account of its own product, and the figures are as its customers reported them.
Show stakeholders proposed versions of a feature as clickable prototypes kept side by side, instead of describing each version in a document.
Build stakeholder prototypes from the same coded components the developers use, so reviewers see how the design will behave in the shipped product.
Derived by Working Surface from the article; more in the Playbook. Source line: The agent did the editing, so I could focus on the concepts behind each tier.
25 October 2026: AI agents and people outside engineering can change real products safely once a team has built shared templates, written rules and automated checks.