I build database engines from the ground up. As the primary creator and core contributor behind three major open-source data infrastructure projects, I focus on solving the hardest challenges in vector search, hyper-convergence, and memory-centric storage.
-
๐ญ Currently Building (Core Kernel):
-
๐ Created From Scratch:
- Milvus: Founded and built the world's most popular open-source vector database, powering AI applications for thousands of enterprises.
-
๐ง Deep Expertise:
Database Kernel InternalsDistributed ConsensusVector Search AlgorithmsStorage Engine DesignGoRustC++ -
๐ฌ Ask me about: Building a DBMS from zero, scaling vector search to billions, HTAP architecture, or the internals of Milvus/MatrixOne/Memoria.
| Project | Description |
|---|---|
| MatrixOne | Hyper-converged cloud-native database merging OLTP, OLAP, and Stream processing. |
| Memoria | Next-generation memory-centric storage engine for modern workloads. |
| Milvus | The original cloud-native vector database for AI, created from scratch. |





