Occasional technical writing on things I've built, broken, or had to think hard about. No regular schedule — only published when there's something worth saying.
A real account of going from zero to a stable production LLM inference server in under three months, with no prior experience, a limited budget, and no one to ask.
Read article →An honest look at AI coding tools from someone who maintains LLM inference and uses a coding assistant every day. What actually works, what fails quietly.
Read article →