A REFERENCE & SELF-ASSESSMENT INSTRUMENT

The AI-Native Engineering Maturity Model

Ten dimensions for scaling AI

and the decisions that still have to be human.

What it is

AI can now write the code, the tests, and the ticket. What it can't do is decide which of those you actually trust it with — and that decision looks different at every level of maturity.

The model breaks that down across ten dimensions of engineering work. Each one has four stages, from AI used off the books to AI built into how the work gets done. For every dimension you get what the stages actually look like, a question to place your org, and the first concrete move from where you are.

How to use it

Place your org on the four-stage scale, one dimension at a time. Then ask a few other people where they'd put it — the most useful conversation is usually in the disagreements. Focus on the one or two dimensions that matter most right now. Once you've placed yourself, the stage above is your next move.

Get the model

Free, and there's no email wall. This page always links to the current version — future editions land here as the model gets updated.

PDF · 15 pages · v1.0 (July, 2026)

Versions

v1.0 (July, 2026) — current

v2.0 (Target: Summer 2026) — coming soon

Who made this

I'm Ed Schaefer. I work with engineering leaders on scaling AI without losing the judgment that has to stay human — moving from shadow-and-sanctioned use to a practice that's governed, measured, and actually embedded. If you want to run this with your team, or work through a single dimension, that's the kind of thing I do.