Field Notes is here. Hold your applause.

We've built things, broken things, and fixed things. Now, we're writing it down.

That is to say: we have a new engineering blog called Field Notes. Naming it took over a week, because there are two hard problems in computer science: naming things, cache invalidation, and off-by-one errors. Finally, we found a name that nobody objected to, which is the highest praise a naming committee can earn. Field Notes it is!

Why another blog?

Most company blogs are at best two parts filler to one part useful, with headlines like "Seven Trends Shaping the Future of Everything." We read them too, but we lose interest somewhere around the next new way to solve everything.

The good stuff lives on personal blogs and in feed readers, written by someone who hit a wall on Tuesday and wrote it up on Wednesday. It's a little rough, but it's honest, and that's the bar for Field Notes.

Build to learn

We've spent a year and a half building with AI. We've learned a lot.

Most public writing about this stops at the demo. Demos are great, but most never ship. The lucky ones that do meet the same fate: someone paged at 2 a.m., wading through AI slop. We want to write about what we learn from building, whether it becomes a blockbuster success or a... moral victory... with learning attached.

We're a build-to-learn culture, and engineers tell us it's a big part of what keeps them happy here.

We launched an internal Builders Friday: a regularly scheduled day to pick up a non-core pet project or scratch an itch you've been meaning to fix. Whatever comes out of it gets shown at our standing Demofest. An innovate-in-the-open Slack channel gives the hackier experiments somewhere to land.

Separately, like many companies, we run an annual company-wide hackathon: a full week to work on pie-in-the-sky greenfield projects.

Field Notes is our attempt to take that culture, all the highs and lows, and make them public.

What to expect

• Things we built, and why. Including the parts we'd do differently.
• Things that broke. Postmortem energy, no blame.
• Experiments with real data. Including the ones that came back "meh."
• Code you can steal. Interactive visuals when they help, not when they're cute.
• Follow-ups. Everyone focuses on "how it started." We think "how it's going" is worth sharing too.

Every post says what we tried, what worked, and what we're not sure about. Our Muse post admitted its sample was tiny and untested on real traffic. We weren't kidding!

Field Notes is by engineers, for engineers. No thought leaders, no clickbait, and nobody who thinks "load-bearing" is a marketing term. If we wouldn't send it to a coworker in Slack, we won't publish it. That's our promise.

Who writes this?

Every post comes from the people who built the thing, not from someone who interviewed them for twenty minutes and guessed the rest. If you read a post about a migration, the author is the one who got paged during it.

---

More posts are in the queue, so add us to your feed reader and hit that subscribe button. There's no like button, and therefore we won't ask you to smash it.

That's the intro, and it compiled on the first try, which will never happen again. LGTM 🚀