Kevin L.

About Kevin L.

I am interested in the practical applications of machine learning and deep learning to solve the world\'s problems.

Experience

1. Senior Data Scientist, Ai

   Meta

   Aug 2023 — Present · Seattle, WA, US

   Architected and deployed \"KevinBot,\" a production agentic SQL assistant (RAG, intent routing, skill orchestration) with strict source-of-truth whitelisting, enabling self-serve analytics and cutting ad-hoc DS ticket volume by 40%.• Engineered social-graph-based targeting algorithms (routing dense graphs to friend tiles and sparse graphs to discovery banners) to solve severe cold-start constraints, driving Horizon Worlds penetration from 1% to 10% and doubling Community DAU QoQ.• Defined the core PMF metric framework and north-star telemetry for Communities, shifting organizational focus to Play/Travel entry points and securing premium experimental feed placements.• Developed a Bayesian causal inference framework leveraging hierarchical pooling and CUPED variance reduction to combat chronic low-sample underpowering, making 80% more experiments decisionable and reducing analysis time by 35%.• Shipped network-level fraud and spam detection models built on stratified ground truth sampling and tiered enforcement policies, slashing scam-message visibility by 30–50% across social surfaces.

Education

• University of California, Davis

  Bachelor's Degree, Economics

  2010 — 2014

• Homestead High School

  High School Diploma

  2006 — 2010