Blog

Engineering Notes

Lessons from building production AI systems on constrained hardware. No theory — just what worked, what didn't, and why.

Career
Coming soon

From Program Manager to Production ML

What 22 years of enterprise IT taught me about building AI systems — and what it didn't.

Infrastructure
Coming soon

Running Production ML on Consumer Hardware

How a single GPU server replaced $18K/month in cloud AI inference — and the tradeoffs that come with it.

Engineering
Coming soon

What 400 AI Podcast Episodes Taught Me About LLM Context Management

The phased generation pipeline that solved a 104K token overflow problem.

First posts coming April 2026. In the meantime, see what I've built or read my story.