Minnesota AI
A field lab for AI discoverability engineering.
Minnesota AI runs controlled experiments on how AI systems find, parse, cite, and route users to websites. Every finding is published openly, backed by live data from real sites with real traffic.
David Valencia
Analyst turned systems builder. I spent years working with raw data — finding patterns, building automations around recurring signals, and turning messy inputs into reliable outputs. That work ran through analytics, data pipelines, API integrations, and workflow systems before AI was part of the conversation.
When AI changed how people find information, I recognized the same infrastructure patterns I had been building around for years. The signals were familiar. The systems problems were familiar. The technology was new.
I founded Minnesota AI in 2020. Now I run controlled experiments on how AI systems find, parse, cite, and route users to websites. I build the sites, operate them with real traffic, measure what happens, and publish the data on the live feed.
The Research
Three questions drive the work:
What can AI access? — How semantic structure, markup, and machine-readable patterns determine whether a model can parse a site at all.
What does AI understand? — How content clarity and consistency determine whether a model builds an accurate representation of what an organization does.
What does AI choose? — What makes a model confident enough to cite, recommend, and route users to a specific site.
Each question has a dedicated research area. Findings are published as articles, and live crawler and traffic data is available on the live data feed.