Manage data quality issues
The outcome: manages data quality issues by impact through to verified closure.
Builds on: Analyse the root causes.
Which AI model is best at manage data quality issues?
| Model | Provider | Tier | Test score | Knobits mastered (expected) | Cost per test run |
|---|---|---|---|---|---|
| gemini-4-argon | Flagship | – | 84% | $0.021 | |
| gemini-3.5-flash-lite | Low cost | – | 55% | $0.003 | |
| gpt-6-luna | OpenAI | Low cost | – | 81% | $0.002 |
| claude-haiku-5-5 | Anthropic | Low cost | – | 85% | $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 9 knobits of this activity
5 that AI does well, 0 it does partly, and 4 where people still lead.
| ID | Knobit | Type | AI exposure (models mastering) |
|---|---|---|---|
| Q-601 | Recognises a data quality issue severity set by the reporter | Skill | AI does this well 3/4 |
| Q-602 | Recognises a data quality issue owner unable to fix its cause | Skill | People still lead 0/4 |
| Q-603 | Recognises a data quality issue closed with no verification | Skill | People still lead 0/4 |
| Q-604 | Recognises duplicate data quality issues left unlinked | Skill | People still lead 0/4 |
| Q-605 | Recognises an issue on a critical data element with no impact stated | Skill | AI does this well 3/4 |
| Q-606 | Recognises a data quality issue past its target with no escalation | Skill | AI does this well 3/4 |
| Q-651 | Explains impact-based severity of data quality issues | Knowledge | AI does this well 4/4 |
| Q-652 | Explains the data quality issue lifecycle | Knowledge | People still lead 0/4 |
| Q-653 | Explains resolution targets and escalation of data quality issues | Knowledge | AI does this well 3/4 |