Why I built it
I formulate perfume, and the tool I relied on didn't fit how I actually work. So I wrote the requirements, designed the system and built the replacement: a raw-materials library, a formula editor with live percentages and cost, dilutions, and iteration notes that stay visible as a formula evolves.
How AI changed the work
- Requirements first. I wrote structured requirements and marked every assumption for review, the same discipline I ask of my team.
- AI as engineering partner. I directed the architecture, design system and features; AI wrote much of the code under rules I set.
- Real release practice. Preview and production environments, review before release, and a beta program with signup and feedback.
Quality bar
- Accessibility: WCAG 2.2 AA in light and dark mode, with an automated contrast audit on every design token.
- Privacy by design: users' formulas live only in storage they own, never readable by me.
- Safety by design: unknown safety data is never treated as safe.
I don't just manage AI-assisted work. I ship it.
Why it matters for leadership
Building a real product with AI taught me where it accelerates work, where it needs guardrails, and what a team has to change to use it well. That's the experience I bring to leading AI-native design teams.