GeoReady: an open engine, a public method, and a book.
GeoReady is built independently in Italy by Juan Camilo Auriti. The audit engine it runs is
open source, and the method behind it is written down in full.
What I built
An engine you can run yourself, and an app that runs it for you.
GEO Optimizer
MIT · Python · PyPI
The open-source audit engine. It measures how ready a site is for AI search across
9 scoring categories, with
16 CLI commands to audit, fix and monitor. The weights are
published in models/config.py.
The hosted app built on the same engine. It runs the audit on a schedule, keeps the score
history, and alerts you when a change drops a site's score. It is built for agencies, SEO
freelancers and developers who look after more than one site.
The problem I set out to fix is the black-box score: a number between 0 and 100 with no way to
see how it was reached, so no way to tell whether a fix moved it for a real reason. A tool that
criticises that cannot work the same way.
So the scoring is public. Every check, every weight and every band is in the repository, and
you can run the same audit on your own machine and get the same number. If you think a weight
is wrong, open an issue and argue it in the open.
AI visibility infrastructure should be auditable by anyone, not locked behind proprietary black boxes.
arXiv:2510.11438AutoGEO, which builds on the Princeton studyarXiv preprint
Where a check goes beyond these papers, the reasoning is documented on the
research page.
The book
The method, written down.
AI Search Engineering is the long version of what the engine
checks: a technical guide to Generative Engine Optimization, AI citations and agent-ready
websites. The book explains the
reasoning; the engine runs it.