Stripe Develops AI-Powered Prototyping Platform
Design manager Owen Williams at Stripe built an internal AI-powered prototyping platform called Protodash. This tool allows designers and product managers to create high-quality prototypes without writing code.
Protodash uses Cursor rules, React components, and Stripe's design system to provide a web-based prototyping studio that runs in development environments. The platform includes features such as design review modes, variant testing, and AI-powered iteration.
According to Williams, the tool has been widely adopted by both designers and product managers at Stripe, fundamentally changing how they approach prototyping and engineering handoffs.
Williams discussed the challenges of using generic AI tools for design, which can result in "blurple slop" – a term he used to describe the poor quality of designs generated by AI. He also explained how Protodash was designed to address these issues and provide a more effective way to create prototypes.
The tool's architecture combines React router, design system components, and MCP integrations to provide a seamless user experience. Williams also highlighted the benefits of running prototypes on dev boxes, which eliminates the need for local setup and allows designers to focus on creating high-quality designs.
During an interview, Williams demonstrated how Protodash can be used to create complex prototypes, including a Black Friday/Cyber Monday pet store dashboard. He also discussed the importance of building internal tools like Protodash, rather than relying on off-the-shelf solutions.
The adoption of Protodash has had a significant impact at Stripe, with product managers now using the tool as much as designers. This shift has fundamentally changed how the company approaches design reviews and engineering handoffs.
