Building tech that includes everyone.

Partner with Product and Engineering on early-stage bets in Instagram's ranking and recommendation systems. Translate complex behavioral signals from large-scale ML systems into product direction — combining statistical modeling, experimentation, and exploratory analysis to surface what's actually driving outcomes. Build forecasting and monitoring frameworks that give teams early visibility into whether a system is performing as intended.

Product vision and zero-to-one execution at a YC-backed AI startup — translating frontier model capabilities into concrete user workflows, prototyping with LLMs as core infrastructure, and iterating concept to working product through eval-driven feedback loops. Safety-conscious thinking embedded from day one with evaluation harnesses to catch model regressions before they shipped.

Designed structured learning frameworks for 2,000+ students that made abstract technical concepts concrete and actionable. Ran a tight feedback loop between observed learner behavior and instructional design — prioritizing transferable reasoning skills over prescriptive fixes.

Led a developer education initiative enabling 34M+ users to build and deploy applications on Pi. Translated an emerging technology into accessible UX and product workflows, bridging the gap between platform capability and non-technical audiences.

Built Vision Pro accessibility features for blind users — new technology enabling inclusive spatial computing experiences through close collaboration across engineering, design, and product.

Worked on gRPC protocol and low-latency systems infrastructure, contributing to performance-critical backend systems within a large-scale engineering organization.

Technical documentation and metric road-mapping at an AI/ML startup building LLM-driven UI optimization and continuous A/B testing systems.