Question 190
At a multi-team data readiness review, an AI project leader faces this situation: An auditor asks which upstream database revision influenced a contested model prediction but no lineage exists. The decision should distinguish missing evidence from confirmed data suitability. Which next project action directly addresses the concern?
Explanation
Correct Answer
In a multi-team data readiness review, the most direct response is: Record versioned extraction, cleaning and feature-generation lineage for each approved training snapshot. This addresses: Model datasets should retain source and transformation identifiers for reproducibility and audit. Its acceptance evidence is: A traced model input can be linked to an authorized source record and transformation history. The project should also remember: Protect lineage logs from unauthorized access when they reveal sensitive source locations.
Incorrect Answers
Require a rare-event sufficiency assessment and choose a controlled pilot or defer deployment. The appropriate correction for the reported case is: Record versioned extraction, cleaning and feature-generation lineage for each approved training snapshot.
Implement supported change tracking with durable checkpoints and controlled replay. This alternative is not the failing dependency described in the question.
Map transfer paths and obtain appropriate legal and security approval before replication. Its verification is distinct from the reported condition: An auditor asks which upstream database revision influenced a contested model prediction but no lineage exists. The decision should distinguish missing evidence from confirmed data suitability.