Operator lens
I approach AI through production behavior. What matters is what changes outcomes when a real user, a real website, and a real model meet.
About
About Me
I run controlled experiments on how AI systems discover, cite, and recommend websites.
My work sits between research and implementation. I build real sites, observe model behavior, and publish what survives contact with the live web.
I care less about abstract AI commentary than about the mechanics that change outcomes in practice: structure, visibility, discoverability, and the conditions that make a website usable inside AI systems.
Organization
Where the work lives.
This personal site is one part of the work. The experiments, frameworks, and implementation lens extend into Minnesota.AI and AIFDS.org as well.
-
Browse writing
Personal site
Writing, theory, and field notes
This site is the public record of the thinking: foundation definitions, plain-language explanations, and the ideas that connect live experiments into a coherent body of work.
-
Visit Minnesota.AI
Minnesota.AI
The field lab
Through Minnesota.AI, I work on applied research, live systems, and experiments built to measure what AI systems actually do in practice.
-
Visit AIFDS.org
AIFDS.org
The open framework
AIFDS.org is where the framework side of the work lives: standards, schema direction, and the system design layer behind machine-readable websites.
Approach
What guides the work.
The goal is not to speculate about AI from a distance. It is to test how models behave around real websites, real constraints, and real user journeys.
Controlled experiments
I prefer live tests over recycled theory. The point is to publish evidence, not to repeat assumptions that collapse under real conditions.
Plain language
A young field needs clear foundations. I write to make complex model behavior legible enough for founders, operators, and technical teams to act on it.
Focus
What this site covers.
Most of the work here is organized around three foundations that explain how websites become usable, trustworthy, and recommendable inside AI systems.
Foundation
LLM Structure
The machine-readable architecture that makes a website legible to models in the first place.
Explore LLM StructureFoundation
LLM Visibility
The conditions that let an AI system access, parse, understand, and verify what a website is saying.
Explore LLM VisibilityFoundation
LLM Discoverability
The recommendation layer: the moment a model decides a website is the right destination for a user ready to act.
Explore LLM Discoverability