AWS Certified AI Practitioner (AIF-C01) · Objective 1.3
Checkpointing long AI jobs
1.3Standard
Embedding thousands of documents takes hours.
Resume after failure.
Choose two designs.
Hint 1: Focus on the requirement
Persist versioned progress and failed items, validate permissions and duplicates on resume, and publish partial results atomically.
Hint 2: Key terms
Review long-running job checkpoints.
Checkpoint
A saved state from which a long-running process can resume.
Atomic commit
Making a result visible as one complete operation or not at all.
Partial result
An output covering only a completed subset of the intended workload.
Resume validation
Checks performed before continuing from a saved processing state.
× Incorrect
Save input version, range, completed items, state, and model version, then validate permissions, duplicates, and failures before publishing partial results.
Detailed explanation
A
CorrectSave input version, range, completed count, state, and model version
Work can resume.
B
IncorrectPublish unverified partial results to search
Incomplete evidence is used.
C
CorrectValidate versions, permissions, duplicates, and failures and publish partial results atomically
Inconsistent evidence is avoided.
D
IncorrectDiscard progress and rerun everything
Time and cost increase.
Check official documentation
Compare the following points with the official documentation.
Path & Pass is an independent learning site and is not affiliated with any certification body. Questions are original practice content based on published objectives.