Build cloud native OLTP database system.

Research and prototype secure container that fully encrypts in-memory database systems while maintaining the performance, leveraging hardware memory encryption engine and Intel SGX based virtualization technologies. To get a flavor of the work, here are a few interesting joint researches in this area from database, OS and hardware architecture (secure hardware) communities:

https://petsymposium.org/2019/files/papers/issue3/popets-2019-0052.pdf

https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8418608

http://www.fkerschbaum.org/dbsec17a.pdf

https://eprint.iacr.org/2016/086.pdf

Research and build distributed transaction management system. Distributed lock manager, distributed deadlock detection (distributed 2PL) and their integration with the record access layer, 2PC and consensus (for each partition level replication group).

A side project: graph database and graph processing engine. Graph partition research. Reverse-engineered Titan/Janus graph database internals.

Research and development multi-dimensional range query index over key-value store.

Real-time, distributed and massively parallelized, spatial-temporal database system: built high performance Http/Rest API servers(a.k.a the Front nodes) of SpaceCurve's database cluster, inter-engine communication fabric, data ingestion pipeline, query engine and distributed spatial JOIN operator, replication controller and rebalancer(for data partitions).

http://www.jandrewrogers.com/2015/10/08/spacecurve

http://www.jandrewrogers.com/2015/03/02/geospatial-databases-are-hard/