In-house analyst
Describe a crossing or a question about this procurement data. The analyst answers with what has already been measured, gives the denominator and the unit of each variable, and points out which of this dataset’s known traps apply to your question — before you spend hours on it.
It does not calculate. The analysis engine does, with 290 tests and guards that exist because a week of defects demanded them. A language model adding up billions in its head produces a plausible and wrong number, and plausible and wrong is the worst possible outcome, because nobody checks it. The analyst cites measured numbers; when a question needs a number that does not exist yet, it says so and describes the harvest that would produce it.
Answerable with today’s data?
What this will not prove:
Numbers cited
| Field | Value | Scope |
|---|
Variables, with denominator and unit
| Field | Role in the answer | Denominator | Unit |
|---|
Traps that apply to this question
Harvest that would answer it: