No jargon. If this page needs a glossary, it has failed.
An investment firm owns things for reasons. Those reasons get written down once: in a memo, an IC deck, an email, and then the world moves on. Six months later the position is still there and nobody has systematically checked whether the reasons still hold.
Meanwhile every tool on the desk is monitoring the thing: the price, the news, the filings. None of them is monitoring the reason.
It takes the reasons and makes them first-class objects. A thesis becomes a small set of named assumptions, "hyperscaler capital spending stays above 20% through FY28". Each with the evidence behind it and the condition that would prove it wrong.
Then it watches those assumptions rather than the ticker. When something arrives that bears on one, it investigates: gathers evidence, deliberately looks for the evidence that argues the other way, weighs how much it matters given your exposure, and only then decides whether to interrupt anyone.
Most mornings, almost nothing. That is the point. Of three hundred positions, perhaps three have something worth your attention, and the system tells you plainly that it checked the other two hundred and ninety-seven and found them quiet.
You accept, investigate, dismiss, or change the thesis. Whichever you choose is recorded permanently: what was proposed, what you decided, what evidence existed at the time. Nothing is overwritten.
Later, when the outcome is known, that record is what lets the firm learn something real about itself: which assumptions it habitually gets wrong, which evidence it underweights, how early it usually sees a problem.