Decisions, frozen in their moment

The data as it was, the rules as they were

Each decision in ADIP produces a point-in-time record, and the record is immutable: nothing recomputes it, nothing restates it, nothing quietly updates it under a later configuration. What was true when the call was made stays true on the page.

Most systems can archive the data. Almost none preserve the rules — yet rules drift, and a number severed from the rule that produced it no longer means what it meant. So the record freezes both, together:

• The data as it stood — orders, positions, promise dates, the signal that opened the decision.
• The rules in force — revenue treatment, credit policy, thresholds, as of that month, not as of today.
• The alternatives that were priced, with the constraints and preferences applied to them.
• The decision itself, who made it, and the rationale — the reason for the call, kept with the call.

Sit in the pilot seat

Judging a landing from the tower replay is easy — the tower already knows how it ends. ADIP's records put the reviewer in the pilot seat instead: open a decision from last March and you see what they saw. The same numbers, the same rules, the same options at the same prices. The page remembers the moment, not the aftermath.

That changes the review conversation. The question stops being “why does this look wrong in today's numbers?” and becomes “given what was knowable, was this the right call?” — the only version of the question that improves anything.

The end of cheap second-guessing

Today, second-guessing costs nothing and answering it costs an investigation. Reconstructing what a planner actually knew in a given week means correlating logins, system logs, and database change history — a litigation-grade effort nobody mounts for an ordinary decision. So hindsight wins by default, and people learn to decide defensively.

A frozen record reverses the economics. What was known, what was priced, what was chosen and why — it's all on the decision's page, retrievable in seconds. Hindsight is welcome to disagree. Now it has to argue with the record.

Improvement gets honest

None of this shields people from scrutiny — it makes scrutiny worth having. With the moment preserved, you can finally separate a bad outcome from a bad decision, and a bad decision from a signal that arrived too late to matter.

Those are three different problems with three different fixes: better judgment, better rules, or earlier sensing. The record tells you which one you have — and the loop begins again where every decision begins, at the signal.