Stripe's Protodash: How AI prototyping tools are reshaping product design

The Gist

• Stripe built Protodash, an internal AI tool for rapid prototyping using their design system
• Protodash evolved from Cursor rules and React components into a full web-based platform
• PMs became unexpected power users, shifting from memos to working demos
• Generic AI design tools often fail to capture specific design system nuances

• Protodash, Stripe's internal AI prototyping tool, enables designers and PMs to create production-quality prototypes quickly by leveraging Stripe's design system.
• Generic AI prototyping tools often produce subpar results ('blurple slop') because they lack knowledge of specific design systems.
• PMs have become unexpected power users of Protodash, improving collaboration with designers and enabling faster idea validation.
• Protodash evolved from a simple set of Cursor rules and React components into a full-fledged web-based prototyping platform.
• Internal tools like Protodash are most effective when tailored to a company's specific culture and workflows.
• Owen Williams advocates for being specific in AI prompts and not hesitating to start over when results are unsatisfactory.

• Invest in internal AI tools that align with your company's design system and workflows to avoid generic, low-quality outputs.
• Encourage PMs to use prototyping tools to unblock themselves and validate ideas earlier, improving collaboration with designers.
• Be specific in AI prompts and don't hesitate to reset the context window if initial results are unsatisfactory.

• half a million dollars (Gross volume of a dashboard prototype created using Protodash.)