Kazuki Ohta
Founder at Treasure Data, Inc.
San Francisco, California, United States
Summary
🤩
RockstarKazuki Ohta is a founder and seasoned software engineer based in San Francisco with 16 years of experience building backend systems and developer-focused infrastructure. As an early leader at Treasure Data (CEO & Co-Founder, former CTO) he blends product leadership with deep hands-on engineering across distributed systems, logging, and serialization. His open-source contributions reveal sustained expertise in high-performance messaging and logging—implementing Java MessagePack-RPC, expanding MessagePack's Java test suite, and improving Fluentd and its Python logger. He comfortably spans backend development, QA/test automation, and DevOps, with practical improvements from transport layers to build scripts and memory-aware process management. Notably, his work includes adding asynchronous futures and robust timeout handling to RPC code and introducing Unix domain socket support and timestamping to a production logging client. He brings entrepreneurial pragmatism and a track record of shipping resilient, production-grade tooling used in large-scale data and logging ecosystems.
16 years of coding experience
Github Skills (46)
unit-testing10
msgpack10
netty10
python10
testing10
data-serialization10
java10
serialization10
ruby10
javas10
junit10
performance-optimization10
unicorn10
logging10
fluentd10
Programming languages (12)
Github contributions (5)
kzk/unicorn-worker-killer
Nov 2012 - Apr 2021
Automatically restart Unicorn workers based on 1) max number of requests and 2) max memory
Role in this project:
Backend Developer
Backend DeveloperContributions:44 commits, 7 PRs, 9 pushes in 8 years 6 months
Contributions summary:Kazuki primarily contributed to the core logic of the `unicorn-worker-killer` project. Their work focused on bug fixes, particularly concerning the accuracy of memory usage calculations and variable scopes. They refactored the codebase, changing namespaces and introducing features for disabling certain functionalities. The user also implemented changes to improve error messaging and compatibility across different Ruby versions.
memoryunicornpythonworkers
fluent/fluent-logger-python
Oct 2011 - May 2015
A structured logger for Fluentd (Python)
Role in this project:Back-end Developer
Back-end DeveloperContributions:46 commits, 7 PRs, 7 pushes in 3 years 7 months
Contributions summary:Kazuki primarily contributed to the development of a Python-based logging library for Fluentd. They implemented core functionalities such as message formatting, sending events, and handling connections to Fluentd. The user also introduced features including timestamp support, and a Unix Domain Socket sender, while refactoring the sender module. These changes improved the library's functionality and connection methods. The user also created a handler class for integration with standard Python logging.
pythonstructuredloggerfluentdlogging
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.