Supporting Facebook Notification ML and Product Experiences, Content Ecosystem, and Facebook Newsfeed and Notification Ranking Infra, focusing on the growth of Facebook DAU, high quality content production, and ML infra foundation.

Supported the Facebook Community Discovery org, developing series of community discovery surfaces and products (e.g. Facebook Community Tab, Events, Hashtags and Trending Explore) to connect everyone to the right content and the right people so they can build, engage and strengthen communities that matter to them, and therefore ensure Facebook’s mission to "Uplift the world through safe & inclusive communities".Supported the Facebook Search org, building search experience within Facebook App, and developing search and recommendation infra platforms for products across Meta's family of apps.Supported Facebook Groups growth, building Groups recommendations and Groups invite/share experiences.

Built up and supported Facebook App Commerce Intelligence and Quality org, working on product ranking and recommendations, product search, commerce ranking infra for Facebook Marketplace, Facebook Shops and Facebook Jobs. Also opened up new areas and built up the Marketplace Integrity team for scam and policy violation detection, Product Intelligence team for product catalog and content understanding with computer vision and NLP, and Marketplace Ads ranking team.

Supported Marketplace Intelligence team, working on personalized product recommendation ranking, ranking data and system infra, quality and integrity framework, and leveraging computer vision and natural language processing technologies for product understanding.

Game Ranking:• Lead the design and build of personalized game ranking and recommendation system;• Optimized social interaction channels for Facebook game users.Ads Strategic Metrics:• Designed and built data mining and visualization system to monitor and diagnose Facebook ads system health;• Built supply and demand analysis framework for Facebook ads ecosystem.

Worked on machine learning for natural language and speech processing, including:• Twitter user interest extraction• Active learning for sentiment analysis over movie and product reviews• Social algorithms for extracting lexicon semantic similarity over Wiktionary • Language model adaptation for large vocabulary continuous speech recognition (LVCSR)

• Designed and implemented an intelligent query understanding system for local business search• Implemented data mining algorithms over a large scale map-reduce system

• Designed and implemented paraphrase detection algorithms for voice search based SMS message reply system in automobiles

• Designed and implemented dynamic and static topic language model adaptation for SRI Mandarin speech recognition system for DARPA GALE project