About this Session
In large organizations, small differences add up. Teams solve the same problems in slightly different ways: different constraints, budgets, skill sets, vendors. This leads to tech debt and a fragmented experience for users. And AI may just make things worse by speeding up that fragmentation.
AI doesn't know what's _right_. It only knows what's _likely_. Probability isn't enough when you're building for millions of people who need consistent, reliable access to services like healthcare, housing, or help.
But a well-structured design system gives AI something it can't get from training data: your design and engineering decisions, encoded as infrastructure. Instead of making slop design and engineering decisions, AI relies on your guidance to wire together pieces you've already built and tested.
We'll look at real examples from the New York State Design System, like quickly turning a 5-page PDF form into an accessible web application. And we'll go through how Figma, Code Connect, and MCP servers provide layers of guidance that keep designers in the loop while giving AI guardrails that scale.
This isn't about limiting creativity. It's about coordinating work across dozens of agencies to create a consistent experience for the people depending on it (and for the people who have to build and maintain it).
