Tech Lead for Facebook Monetization, responsible for designing and building end-to-end ad retrieval and ranking systems for new and emerging ad surfaces

Play Apps Personalization & Recommendation

Manage multiple machine learning teams to help 2B+ users discover great apps and games at the Play Store. Responsible for game recommendations, user understanding, ML infra and in-product surveys, etc.

YouTube TV Search & Discovery

Lead Home Page Recommendation and Ranking Signals of YouTube TV. Built the first deep neural network based recommender system for YouTube TV and various ranking signals (e.g. relatedness, predicted viewership, predicted genres, etc), collectively increasing Home Page watch time by ~20%.

Image Search

Built the Guided Search feature on Image Search as IC.

Worked as part of Amazon Search & Discovery Group, worked on building the Catalog Backfill System to automatically discover and extract valuable product facets from unstructured text data (e.g. product title and description) using a combination of information retrieval, natural language processing, and machine learning techniques and write them back into structured catalog fields (e.g. display size).

Worked as a software developer in Oracle Fusion GRC team in Irvine. Mainly focused on building next generation user interface for Oracle Governance, Risk and Compliance (GRC), one of the Oracle’s Financial Management Solutions, using Java, javascript, and Oracle Application Development Framework(Oracle ADF).

Worked as a research programmer in the Computational Behavior Group(CBG) involving in developing data-driven computational models to help understand human behavior and improving decision-making in dynamic and uncertain domains. Mainly worked on building a multi-threaded, goal-based AI for Starcraft, a real time strategy game developed by Blizzard using modern artificial intelligence and machine learning methodologies.