Lei Xu
Co-Founder CTO at LanceDB (YC W22)
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolLei Xu is a Co-Founder and CTO with 11 years of experience building scalable data and AI infrastructure from San Francisco, currently leading LanceDB (YC W22) while serving as an Apache Hadoop committer and PMC member. His background spans deep distributed-systems work at Cloudera and Cruise—shipping production HDFS features, erasure coding, and data processing pipelines—and research credentials with a PhD in Computer Science. He blends hands-on engineering and leadership, steering teams through complex storage and consistency challenges and turning research ideas into production-grade systems. An active open-source contributor, he has improved ML tooling such as CTGAN for synthetic tabular data and drives developer-facing database tooling at LanceDB. Notably, his work bridges low-level storage internals and emerging AI retrieval stacks, reflecting a rare combination of systems rigor and machine learning curiosity.
11 years of coding experience
7 years of employment as a software developer
Bachelor of Engineering Electronic Science and Technology, Bachelor of Engineering Electronic Science and Technology at Wuhan UniversityDoctor of Philosophy (Ph.D.) Computer Science, Doctor of Philosophy (Ph.D.) Computer Science at University of Nebraska-Lincoln
English, Chinese
Github Skills (14)
data-generation10
pytorch10
machine-learning10
datatable10
python10
generative-adversarial-network10
tabular10
command-line-interface8
cli8
command-line8
data-science8
pandas6
scikit-learn6
scikit6
Programming languages (3)
Github contributions (5)
sdv-dev/CTGAN
Sep 2019 - Sep 2020
Conditional GAN for generating synthetic tabular data.
Role in this project:
ML Engineer
ML EngineerContributions:30 commits, 9 PRs, 37 pushes in 1 year
Contributions summary:Lei's contributions primarily focused on enhancing the CTGAN framework for generating synthetic tabular data. This involved implementing command-line interface (CLI) features, adding functionality for saving and loading models, and integrating conditional sampling capabilities. The user also addressed pip installation issues, added example datasets, and made improvements to the setup configurations. These changes collectively streamlined the model's usability and functionality.
pytorchdeep-learningconditionaltabular-datagenerative-adversarial-network
leix28/leix28.github.io
Dec 2016 - Jan 2025
Contributions:3 PRs, 66 pushes, 1 branch in 8 years 2 months
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.