The cockpit keeps growing. Cluster, center stack, passenger display, head-up display, and rear seat screens all have to feel like one product, and they have to change faster than a three-year program cycle allows. Most teams answer this with a different toolchain per screen and a design system that drifts the moment it leaves Figma.
This session shows a different answer, unified under AGL. We designed a new modern IVI experience in Figma, exported its tokens as the contract between design and code, built it once in Flutter, and ran it across multiple displays and form factors from a single codebase. Then we pushed on what comes next: generative UI that assembles a surface around what the driver is actually trying to do, personalization that reshapes the cockpit per profile, and 3D content living in the same render tree as standard widgets.
We will walk through the architecture, the testing infrastructure, and the iterative development process that closes the gap between Figma, Flutter, and agentic engineering, while also highlighting where the AGL Flutter toolchain still has gaps worth filling as a community.
