Skilirium

Recognises a past event judged against present data

K-715 · Skill knobit · Activity: Derive data quality rules · Competence: Data modelling

A rule about a past event checked against today's state: today's reference data, current_date, the current version of a dimension. Rows that were valid when they were recorded start failing when the world moves on, and the failure count grows without a single bad row being loaded.

Can AI do this?

AI does this partly Some of the tested AI models master this; others do not show it. 6 of 10 models master it.

ModelTierProbability of masteryStatus
gemini-4-argonFlagship 83%Not shown
gemini-3.1-pro-previewFlagship 100%Mastered
gpt-6-astraFlagship 100%Mastered
claude-fable-5-1Flagship 100%Mastered
gemini-3.8-flashMid tier 100%Mastered
gpt-6.1-solMid tier 99%Mastered
claude-sonnet-5-5Mid tier 100%Mastered
gemini-3.5-flash-liteLow cost 14%Not shown
gpt-6-lunaLow cost 67%Not shown
claude-haiku-5-5Low cost 83%Not shown

Mastered means a probability of at least 90% that the model has this knobit, estimated from its graded answers. “Not shown” means the answers do not settle it either way. How the models are tested.

What it looks like in practice

Expected result: The check and the as-at reading it needs

How it is assessed: Flags the deficient case and passes the acceptable one. Deficient: account_currency_code checked against today's ref_currency; after EEK was withdrawn in 2011 every account opened in kroon fails Acceptable: The check joins the ISO 4217 version valid on the account's opening date; current_date only in rules that are about now

Also known as: Judged at the wrong time (data quality rule)

Related knobits

Builds on

All 31 knobits of Derive data quality rules.

Source: Review checklist 07-dq-rules, row 15; ISO 25012; DDL 2014; this skill (06-physical.md)