Jun Kim

I work in global business planning at a major Korean food company, where I've spent the last few years building the data infrastructure my team actually needed - designing pipelines, standardizing cross-market data, and making analytics accessible to people who just need answers.

Outside of work, I build things I wish existed. oMLX is an open-source LLM inference server for Apple Silicon - continuous batching, SSD KV caching, multi-model serving, managed from the macOS menu bar. What started as a personal frustration with existing local LLM tools has grown into a project with 2,500+ GitHub stars.

If you'd like to connect or just say hello, feel free to reach out.