* Lead the team to build end-to-end services of Kuaishou’s recommender systems for a variety of business including short videos, live and e-commerce. * Lead the team to build one of the largest recommendation model online training and serving platform in the world. Support developing large-scale recommendation models such as SIM. Reduce the training and serving cost by leveraging new hardware such as PMem, CXL solutions and NVIDA GPUs.* Lead the team to design and build the next generation recommendation platform. The team designs a python-based core engine on which recommendation algorithms can be easily implemented and deployed in a distributed environment.

Lead the team to upgrade JD’s search ranking system from tree-based system to deep-learning based ranking system, and drive the key metric to increase by xx% YoY.

Instagram Ads Delivery (Targeting, Ranking, CTR prediction)

• Storage engines for high-performance computing and machine learning systems• Metadata management for filesystems• Time-series databases and real-time processing framework

Work on Hadoop MapReduce System

Research problems in large-scale social networks.

Work on NLP problems