About Me

I lead research and applied AI for Trust & Safety at Spotify. My career has been a series of pivots that only make sense in retrospect.

I started as a humanities major, earned a PhD in Epidemiology, spent years in academic research that influenced federal legislation, and then moved into technology — first at Indeed, where I founded the company’s Responsible AI function. While at Indeed, I was a 2-time AI Ethics & Governance Fellow at the Berkman Klein Center, where I co-created AI Blindspot, a framework that helps organizations identify structural blind spots in AI systems.

I joined Spotify in 2021 to build its first Trust & Safety data science team, grew it into a research & insights function, and later led a combined function of research, safety strategy and responsible programming. More recently my work has moved toward generative AI: building the company's AI stress-testing practice, and developing ways to evaluate the safety of conversational AI assistants.

The subjects have changed a lot. What I've enjoyed most in each place hasn't: taking on a daunting problem, getting scrappy, and figuring out a way to tackle it.