Liqi Geng is a backend engineer with 8 years of experience specializing in distributed storage and database internals, currently contributing to the TiKV Storage Team at PingCAP in Shanghai. He has hands-on expertise in TiDB/TiFlash/TiKV, implementing SQL function pushdowns, improving MPP behavior, and hardening Raft, region merge, and snapshot paths to boost consistency and performance. His contributions include production-facing fixes like decimal rounding, null-aware semi-join handling, and non-blocking MPMC queue improvements—work that directly impacts TiDB Cloud’s analytical engine used at scale. Before PingCAP he built backend systems for large-scale mobile games, bringing operational rigor and performance focus to low-latency workloads. Trained in information security at Hunan University, he combines systems-level correctness with practical engineering for reliable distributed databases. A detail-oriented open-source contributor, he often tackles subtle build and concurrency issues that improve long-term maintainability.
Distributed transactional key-value database, originally created to complement TiDB
Role in this project:
Back-end Developer
Contributions:3 releases, 609 reviews, 99 commits in 3 years 1 month
Contributions summary:Liqi primarily contributed to the `tikv/tikv` repository by fixing bugs and improving the stability of the distributed key-value database. Their work involved modifying core components related to Raft, region merging, and applying snapshot tasks. The user also addressed issues related to the processing of read index requests, demonstrating a focus on enhancing data consistency and reliability within the system.
The analytical engine for TiDB and TiDB Cloud. Try free: https://tidbcloud.com/free-trial
Role in this project:
Backend Developer
Contributions:486 reviews, 12 commits, 61 PRs in 10 months
Contributions summary:Liqi primarily contributed to the TiFlash analytical engine for TiDB. Their work involved implementing and supporting new database functions such as `toDayOfWeek`, `toDayOfYear`, and `weekofyear`. The user also addressed issues related to build processes, such as correcting cached entries in the CMake configuration, and made improvements to internal queue implementations. Furthermore, the user actively worked on improving the MPP (Massively Parallel Processing) functionality, including non-blocking functions for MPMCQueue and associated dependencies.
analyticstidbbig-datadistributed-databasetikv
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.