Review the data model against the scorecard
The outcome: reviews a data model against the Data Model Scorecard, with evidence, an independent reviewer and a decision that follows the score.
Builds on: Design the physical data model, Plan the modelling and set its standards.
Which AI model is best at review the data model against the scorecard?
| Model | Provider | Tier | Test score | Knobits mastered (expected) | Cost per test run |
|---|---|---|---|---|---|
| gemini-4-argon | Flagship | – | 94% | $0.021 | |
| gemini-3.5-flash-lite | Low cost | – | 60% | $0.003 | |
| gpt-6-luna | OpenAI | Low cost | – | 95% | $0.002 |
| claude-haiku-5-5 | Anthropic | Low cost | – | 81% | $0.006 |
Test score: mean rubric score on the test tasks for this activity. Knobits mastered: expected share of the activity's knobits the model masters. Cost: mean API cost of one test run at list prices, half price for answers sent as a batch. Method and caveats.
The 8 knobits of this activity
2 that AI does well, 2 it does partly, and 4 where people still lead.
| ID | Knobit | Type | AI exposure (models mastering) |
|---|---|---|---|
| K-1001 | Recognises an incomplete data model scorecard | Skill | AI does this well 3/4 |
| K-1002 | Recognises a scorecard score with no evidence | Skill | People still lead 1/4 |
| K-1003 | Recognises a model reviewed by its modeller | Skill | People still lead 1/4 |
| K-1004 | Recognises a review decision contradicting its total | Skill | AI does this partly 2/4 |
| K-1005 | Recognises a review finding with no owner or action | Skill | People still lead 1/4 |
| K-1006 | Recognises a scorecard score contradicting the validator | Skill | People still lead 0/4 |
| K-1007 | Recognises a review of a version other than the one released | Skill | AI does this partly 2/4 |
| K-1031 | Scores a data model against the Data Model Scorecard | Skill | AI does this well 4/4 |