I've spent 10+ years designing conversations, which sounds simple until you realize how many ways a conversation can go wrong. My work spans automotive, healthcare, and IoT, across multiple languages and markets, and that breadth shaped how I think about systems that need to behave consistently no matter where someone shows up. These days I design for AI, specifically the rules, logic, and constraints that govern how a model behaves across an entire product. Less "script the experience or design a flow," more "define the rules and logic so the model handles the variation." What makes AI genuinely interesting is that good design decisions don't just fix one interaction anymore. The reasoning behind why something works gets built into the model itself, and suddenly that thinking scales.