Hello, World

Every technical blog starts the same way, so let’s get it over with.

Why I’m writing

I’ve been meaning to start a blog for years. The usual friction: finding time, dreading the blank page, convincing myself I don’t have anything worth saying.

What changed my mind is noticing how much I learn when I try to explain something clearly. Writing forces a kind of honesty that internal notes don’t. You can’t hand-wave past the parts you don’t actually understand.

So the primary audience here is past-me, a year ago. Future-me will benefit too.

What to expect

Roughly one post a week. Topics will stay close to what I’m actually working on:

• Machine learning systems — training infrastructure, evaluation, deployment
• AI agents — tool use, multi-agent coordination, reasoning loops
• Software engineering — the unsexy fundamentals that actually determine whether things work

I won’t try to be comprehensive or authoritative. These are working notes from someone in the middle of things, not tutorials from someone who has it figured out.

Subscribe

If you want to follow along, the best way is RSS — no email list, no algorithm, just posts in your reader.

See you next week.