How do design leaders scale quality when velocity is a primary business requirement? At Cloudflare, the answer was to empower designers to become builders, using AI not as a shortcut but as a strategic tool for process architecture and long-term quality. This workshop provides a roadmap for transitioning from manual execution to AI-accelerated strategy, moving well beyond the hype to show what sustainable, thoughtful automation actually looks like in a high-output design organization.

The session is structured around a four-month transition case study: the real journey from AI anxiety to designers contributing directly to production-ready codebases. Participants work through a strategic friction audit using a Sludge vs Soul matrix to identify which tasks are fit for automation and which must stay human. From there, the workshop moves into building a multi-LLM prompt library, architecting platform-agnostic agent instructions compatible with Gemini, Claude, and GPT-4 that scale across teams without locking into a single tool. The final sections address measurement and leadership buy-in, shifting the conversation from time saved to metrics that matter: engineering trust, design system integrity, and UI consistency at scale.

The through-line is a question every design leader working in high-velocity environments eventually has to answer: where is the threshold of good enough, and how do you automate the sludge without losing the soul?