ESSAYS [17] | VOL II | UPDATED 04.24.2026

A letter from the editor · Vol. II opening

On going quiet, and on coming back.

Why this newsletter went dark in 2023, what changed in the year I spent watching, and the bet Vol. II is making on data fundamentals in an agentic world.

I published the last Sanity Check in October 2023. Then I went quiet for two and a half years.

Not because I ran out of things to say — but because the ground was shifting under data engineering faster than I could write with conviction. dbt was in the middle of an identity crisis. Every vendor on earth had rebranded as “AI-powered.” LinkedIn was drowning in hot takes about whether analytics engineering was dead, alive, or just resting.

I had opinions too. I just wasn’t confident I wouldn’t look like a fool in six months. The noise was deafening and the signal was thin. So I stepped back, kept consulting, and watched.

This is the note that explains what I saw, what I’m starting again, and who it’s for.

What changed in the silence

Agentic AI isn’t another rebrand. It’s actually changing who consumes your data — and I mean “who” loosely, because increasingly it isn’t a person at all. Your pipeline’s newest customer is a model making decisions at 3am, not a VP pulling up a dashboard at 9.

When a human misreads a chart, they Slack you a question. When an agent misreads your data, it makes a bad decision and nobody notices until the damage shows up two weeks downstream.

The frameworks are new. The work isn’t. That’s both reassuring and a little terrifying.

I keep reading about “context engineering” like it’s some brand-new discipline — designing data systems that embed rich, machine-readable meaning alongside the data itself. And every time I think the same thing: didn’t we used to call this data modeling? The people pitching the new vocabulary aren’t wrong. They’re renaming work the data community has been doing for fifteen years, with one important update — the downstream consumer changed.

What Sanity Check is, in Volume II

One original essay per week. Practitioner takes on data engineering and analytics through the lens of an agentic world. Grounded in real client engagements at Ray Data Co — the day job is the content engine. Real teams, real broken pipelines, real “what’s our AI strategy?” conversations with executives who read three blog posts and want a 90-day plan.

Sanity Check exists because someone has to take a popular claim about data, pressure-test it against what’s actually happening on the ground, and report back. Yes, no, or it depends. That’s the job.

The recipes are still worth sharing. The kitchen just got a lot more interesting.

Who it’s for

Data leads and analytics engineers at mid-market and enterprise companies. People who own pipelines, models, and stakeholder relationships. Practitioners first, managers second. The reader who runs a team of two-to-ten, reports to a VP or CTO, and is being asked to deliver an AI strategy without the cover or the time to sort the hype from the substance.

You’ll get one essay per week. Frameworks you can use Monday morning, not thought leadership you can’t operationalize. No “AI will replace you” panic pieces. No “just add an agent to it” magic. Practical, opinionated, and honest about what’s working and what isn’t.

The bet

The teams that win the next two years won’t be the ones with the fanciest agent frameworks. They’ll be the ones who got the fundamentals right first — clean models, reliable contracts, documentation a machine can actually parse — and then let agents amplify that work.

Fundamentals first. Agents second. That’s the through-line of every essay in Volume II. If your pipeline breaks every Thursday, I promise you — no agent is going to fix that.

How to follow

  • Subscribe by email — the form below puts the next issue in your inbox Sunday.
  • RSS — the cleanest way in for reader apps.
  • JSON Feed — same content, modern format.
  • The archive — every Vol. I essay plus the Vol. II run.

If you’ve read this far

Drop your email below. The next issue lands Sunday morning, Eastern time. One essay. No drip campaign, no upsell, no “exclusive offer for early subscribers.” If the first one isn’t for you, the unsubscribe link works on the first click.

And if it is for you — reply. Tell me what’s keeping you up at night on your data team. I read every one.


Ben Wilson
Founder, Ray Data Co · Editor, Sanity Check
April 2026