On building this

For five years I haven’t had a personal website. Not for lack of skill — I write backend systems for a living — but because the marginal post wasn’t going to move anything, and the marginal hour was better spent on real work.

That tradeoff has flipped.

The forcing function

I’m moving to Stanford in September. Not for school — my partner is starting in a research lab, and I’m following. The Bay Area is the only AI labor market that matters right now, and I’m headed there with a CV that reads as “backend engineer at an unknown-to-them mid-cap.” That’s not what I am. But it’s what the outside world sees, because nothing about my work is legible from the outside. The git is private. The decisions are private. The scale is private.

The fix isn’t to switch jobs or chase a credential. The fix is to make the work I’ve already done legible — to put real artifacts in the world that say “this person solves the problem you’re hiring for.”

What this site is

Essays. Specifically: essays about systems I’ve built and decisions I’ve had to defend. Things I’ve shipped at production scale that taught me something I didn’t know going in. Email verification. Contact normalization. Credit accounting. The unglamorous parts of distributed systems, which is also the interesting part.

Not tutorials. Not “10 things I learned about React.” Not LinkedIn-coded career posts. The audience is the same audience I want to work with: people who care about the craft.

What it isn’t

Not a brand. Not a funnel. Not a newsletter empire. There is one call-to-action on this entire site, and it’s a link to my LinkedIn at the bottom of the page.

On the design

Anthropic editorial × Bloomberg terminal — academic dark, sharp corners, hairline rules, tabular numerals. The aesthetic does work the prose can’t: it signals seriousness in the half-second before anyone reads a word. I’d rather this site look like a research-lab homepage than a content-marketing landing page, because that’s what the writing here is going to try to be.

If you want the build details, the colophon has them.

What’s next

Posts are queued. The first technical one is on email verification at scale — specifically, the IMAP-timing heuristics we used to infer mailbox existence without sending email — which is probably the single most “interesting backend systems problem” thing I’ve worked on in five years.

I’ll publish it when it’s good, and not before.