Somewhere in the middle of manually cross-referencing public records, I ran into a term I hadn't spent much time with before: open-source intelligence, or OSINT — the practice of gathering and connecting information that is already public, in order to answer a question that no single source answers on its own.
It was a genuine turning point. What I'd been doing out of frustration turned out to be a recognized discipline, with its own methods, its own communities, and its own standards for what counts as a supportable conclusion versus a guess. Investigators, journalists, researchers, and analysts had spent years building techniques for exactly the problem I kept running into: public information that is real, accessible, and still nearly impossible to use.
Learning that discipline changed how I approached the work in three specific ways.
First, corroboration over collection. The value isn't in gathering more documents. It's in confirming that separate, independent sources agree — and being honest about it when they don't.
Second, provenance matters as much as the conclusion. A finding is only as trustworthy as the ability to trace it back to where it came from. Every good OSINT practice keeps the receipts.
Third, connection is the actual work. A pile of public records is not intelligence. Intelligence is what shows up once records are linked by entity, by location, by time, and by relationship — and someone has done the work of checking that the links hold up.
Applying that discipline to my own questions was slow and manual, the same way learning any craft is slow at first. But it worked. Questions that used to take weeks of guessing started taking hours of methodical checking. And once I could reliably do it for one question, in one place, the obvious next thought was impossible to avoid: what if this didn't require a person doing it by hand every time?
That question is what eventually pointed toward the observation that became ALKARTIS.