Zeplin adds AI-driven spec generation that maps components to dev story points

Tech · 4 min read

Zeplin updated its platform with an AI-driven spec generation tool that produces implementation checklists, code snippets, and suggested story points for engineering teams. The model analyzes component structure, interactions, and integration needs to estimate complexity.

Specs include accessibility checks, expected API contracts, and test suggestions, helping teams move from design to planning with fewer meetings. Teams can accept or adjust the suggested story points, and the system learns from historical velocity to refine future estimates.

Product teams saw faster sprint planning cycles, but some engineers caution against overreliance on AI-generated estimations for novel architectures. Zeplin added controls for teams to set their own complexity heuristics and tie estimations to historical data.