JINTAO JIANG

Ads Projects:

• Lead major work in setting up production standards, monitoring and alerts, partial work of business logic, migration procedure and testing for launching a new ads data import pipeline.

Napa (An exabyte scale distributed system data warehouse) Projects:

• Designed and Implemented new mechanism for task scheduling especially around expired data deletion; Helped the team to achieve top priority privacy/compliance and efficiency goals: > 4.5 SWE savings and created/enforced first Deletion Monitoring.
• Investigated and fixed complex internal and customer facing issues with database controller stability and scalability to be able to handle 10x more tables.
• Controller autodrain functionality: Improved accuracy and availability of avoiding unhealthy cells during scheduling. Reduced paging caused by temporary unhealthy servers by 50%.

• Designed and implemented Python Native Support Framework for existing class compiler, method dispatching, and object management architecture using C/C++ and CPython API.
• Identified bottleneck and implemented design in the networking protocol of the database system that achieved more parallelization to allow better scalability while maintaining the consistency and durability of the system.

• Designed and implemented novel indexing schemes in distributed database systems tailored for a new networking protocol (RDMA) to replace conventional Remote Procedure Call (RPC) protocol using learned indexing and B-tree indexing.

• Designed and developed an automatic business interest identification tool for the onboarding process of a web application using natural language processing (NLP) to allow new users to register without entering company information.