Kecy Wu

About Kecy Wu

Economist specializing in causal machine learning and fluent in framing complex business problems into measurable objectives and executable science roadmaps. I focus on building end-to-end scalable causal ML solutions that support high-stakes decision-making and long-term impact.Focus Areas: recommendation systems, causal ML, reinforcement learning, online experimentation

Experience

1. Senior Economist

   Amazon

   Jul 2022 — Present · Seattle, WA, US

   Online recommendation systems:Designing reinforcement learning framework to optimize online content recommendations in cross-team collaborations.Shaped science roadmaps and informed multi-year planning.Developed and integrated hetergenous treatment effect models into online recommendation systems, achieved sizable revenue impact.Designed and deployed end-to-end offline validation workflows using large-scale experiment data, enabling reliable pre-launch evaluation and significantly reducing operational overhead- Marketing bidding optimization:Developed and deployed machine learning models for real-time bidding optimization, enabling faster and more precise decision-making.Designed and executed multi-country experiments to evaluate bidding performance and achieved financial gains- Gen AI / LLM:Prototyped synthetic data validation approaches using Variational Autoencoders (VAE) to assess robustness of causal ML models- Science Community: Reviewer for TKDE/TOIS, Keynote at ICSS, Panelist at NABEPrevious Roles:Economist I, Amazon: 20••••23Economist II, Amazon: 20••••25

Education

• Princeton University

  Doctor of Philosophy - PhD, Economics

  2016 — 2022

• Princeton University

  Bachelor of Engineering (B.Eng.), Operations Research and Financial Engineering

• Hwa Chong Institution, Singapore

  High School Diploma

  2009 — 2011