By the time I'd worked through enough of these questions — professionally and on my own — a pattern had become impossible to unsee. It wasn't that the data was missing. Almost every time, the data existed somewhere, published by someone, sitting in a system that technically counted as "open."
What was missing was the layer that turns a record into an answer.
Each of those is a fragment of a larger system. Understanding only appears once the fragments are connected — across time, across place, across the entities and processes that tie them together. Reading one record in isolation almost never answers the real question. Reading the pattern across many records usually does.
This is where I also had to be honest about the limits of technology alone. I'd spent years around enterprise software and, later, artificial intelligence, and I'd seen both oversold as automatic fixes for exactly this gap. They aren't. A dashboard can display activity without ever explaining its significance. A language model can produce a fluent, confident answer without any of the underlying sourcing that would make it trustworthy. Neither replaces the actual work of connecting facts, checking them against each other, and explaining why a conclusion follows from the evidence.
Put together, the observation was simple to state and much harder to act on: we don't have an information problem. We have an understanding problem. Publishing more data doesn't solve it. Adding another dashboard doesn't solve it. Bolting AI onto an existing portal doesn't solve it either, if the underlying records are still disconnected and the reasoning still can't be explained.
Solving it means building the layer in between — the one that connects records, establishes context, identifies what changed, and explains the evidence behind a conclusion, before handing someone a decision to make. That layer is what ALKARTIS became.