Working Surfacelink · 22 Sep · 04
04
Three figures seen from behind, the backs of their heads showing, work at a tall shared cabinet in a workshop: one slides a newly made drawer straight into it while a small robot hands up tools, the other two test the drawer's runners and lock, and a dusty model cabinet stands abandoned in the corner, the whole scene drawn with wide empty margins on every side.
Bolt Labs

Design engineering at Bolt: shipping components with AI agents

Burak Özdemir · 17 September 2026

Company story · design systems · handoff · Human approval

Read the original
Takeaway

At Bolt, designers now build interface components with AI agents directly in the library that ships, and engineers review those components instead of rebuilding them.

Summary

Bolt is a mobility company whose website serves up to 60 million users a year in more than 50 markets, built on a shared library of over 110 components. Designers used to build working components with an AI agent in a separate internal workspace, and engineers then built each one a second time in the production library. Burak Özdemir, a tech lead at Bolt, describes how the team removed the second build.

Key points
  • The two copies drifted apart. A spacing value changed in one and not the other, and engineers worked from a ticket written before the designers' last changes.
  • Designers now work with an agent in the production library and propose the code change themselves. They approve the visual details, and two engineers review how other code uses the component, its accessibility and its tests.
  • A file of written rules in the library tells the agents to use the existing design values, build on accessible base components and leave unrelated code alone. The agent also checks whether a component already exists before it builds one.
  • For routine components, the time from a working component to a merged one fell from a week to hours, and those hours are review. New patterns and complex behaviour still need engineers during the build.
Implication

A team can split the review of one piece of work by expertise, with designers judging the look and engineers judging the code, against written rules the agent already follows. The account is Bolt's own, and the author says the arrangement suits routine components only.

Source issue

22 September 2026: when agents write code faster than people can check it, teams move the checking into written rules and narrower reviews.