Forward Deployed Design
The practice of embedding designers directly in an organization's operating environment to redesign workflows, authority structures, and services, not just interfaces, in the context of AI transformation.
It is not human-centered design with an AI sticker. It is not service design at scale. It is not management consulting with a design veneer. And it is not a Forward Deployed Engineering role that happens to own Figma files.
FDE embeds technical talent to build, customize, and deploy AI systems. It optimizes the how: model selection, API architecture, inference cost, latency. FDD occupies the gap FDE structurally cannot fill, the why and the whether: the workflow the technology is being inserted into, the authority structures that determine whether it can run autonomously, and the liability architecture that determines whether it will be allowed to make decisions.
When FDE embeds without FDD, the result is the 40% trap: a technical solution that speeds up one step of a broken workflow while the remaining steps consume all the gains. The risk assessors work faster, but the legal department still reviews every case manually. The AI generates recommendations instantly, but the permission structure still requires three signatures. The organization feels busier. The customer waits just as long.
FDD is the companion practice because it asks different questions. FDE asks: how do we make this form easier to fill out? FDD asks: why does this form exist at all, and can an autonomous agent negotiate this data exchange without human intervention?
This is not a criticism of FDE. It is a boundary. And it is where Forward Deployed Design begins.