Led the development of Neural Network based recommendation models for many Google products, including Search, Play, YouTube, etc.

Led the development of an automated Neural Network model generation system for recommendation problems. Successfully deployed in 10+ products.

Led the development of high quality embeddings and the corresponding large scale embedding productionization infrastructure for the most major data sources in Google. Successfully used by 200+ teams.

Led the development and open-source of the first on-device recommendation model solution in the industry.

Worked with Machine Intelligence team on personalized prediction system.

Worked with Ads:Shopping:In-store team on user query local intent project.