Cam Fortin

Delivering the dream I've been chasing for 20 years: near-perfect data that updates itself. I do it by building graphs — not the bar-chart kind, the kind where messy real-world things (companies, people, websites, humans who change jobs) get resolved into one entity that holds its shape. At ZoomInfo I run product for core company, person, and web data: the pipelines that decide what's true about a company and how sure we are. Deterministic rules and agentic pipelines, harmonized into one resolved graph — and every AI prediction ships with explainable provenance or it doesn't ship. Nights, I hack on OnlyData Club: custom datasets with entity resolution, flexible schemas, field-level security, and traceability all the way down, where every enriched field knows which model and which run produced it. On top sits a member graph that stores skills and relationships as claims instead of fields — who said it, how they know, and a human assertion always beats a computed one. The whole thing runs on agent harnesses with graph integrity checks, pre-commit and CI gates, and a Telegram ping the second I need to go check on the AI minions. The Boise AI Collective runs on the same rails, in public. Off the keyboard: mining amethyst and purple agates at my claim in the Owyhee desert, biking conservatively (read: slowly) in the foothills, and trying to be as present as possible while my badass co-founder wife and I watch our daughters grow up way too fast. Ask me about graphs, rocks, or running AI on hardware you own.