PMLE Architecting Low-Code ML Solutions Practice Question

A company is building a document processing pipeline using Document AI to extract data from invoices. They want to ensure high accuracy and handle edge cases where the model may be uncertain. Which THREE steps should they include in their pipeline?

⚠ Common exam trap

PMLE often tests the misconception that a pre-built parser is sufficient for all use cases — candidates overlook that custom training and HITL are required for high accuracy on domain-specific documents.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

Regularly retrain the processor using human-verified data

Option A is correct because regularly retraining the processor with human-verified data continuously improves extraction accuracy and adapts the model to new invoice variations and edge cases over time. Option D is correct because enabling Human-in-the-Loop (HITL) routes documents with low confidence scores to human reviewers, ensuring uncertain or edge-case extractions are validated and corrected before entering downstream systems. Option E is correct because a custom processor trained on the company's specific invoice format captures their unique layouts, fields, and terminology, yielding higher accuracy than a generic model. Option B is not appropriate because using the pre-built invoice parser without modification offers no tuning for the company's specific formats and provides no mechanism for handling uncertainty. Option C is not appropriate because AutoML Vision is an image classification service, not a document entity-extraction tool, and classifying invoice types does not extract the required field data or address low-confidence edge cases.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

✓
Regularly retrain the processor using human-verified data
Why this is correct
Regular retraining with human-verified data directly addresses the accuracy constraint by feeding corrected edge-case extractions back into the processor, letting it learn the specific invoice variations it previously misread. This closes the loop on uncertain predictions, since human review resolves ambiguity that the model alone cannot, progressively raising extraction confidence across the pipeline.
✗
Use the pre-built invoice parser without any modifications
Why it's wrong here
The pre-built invoice parser handles common layouts but cannot be tuned for a company's specific edge cases or uncertain fields, so accuracy stays fixed. It is tempting as a zero-effort start, and would suit straightforward, standardised invoices where no custom extraction or confidence handling is needed.
✗
Use AutoML Vision to classify invoice types
Why it's wrong here
AutoML Vision classifies images, not invoice text fields, so it cannot extract line items or flag uncertain values for review. It is tempting because AutoML trains custom models, and would be correct for sorting scanned document images into categories before a separate extraction step.
✓
Enable Human-in-the-Loop (HITL) to review documents with low confidence scores
Why this is correct
Human-in-the-Loop review directly addresses the stem's requirement to handle edge cases where the model is uncertain. By routing documents whose confidence scores fall below a defined threshold to human reviewers, the pipeline corrects low-certainty extractions before downstream use, preserving accuracy on ambiguous invoices that automated processing alone would misclassify.
✓
Use a custom processor trained on their specific invoice format
Why this is correct
A custom processor trained on the company's own invoice layouts directly addresses the accuracy constraint, since Document AI's pretrained invoice parser generalises poorly to bespoke formats. Fine-tuning on representative samples raises extraction confidence and reduces the uncertain edge cases the stem requires the pipeline to handle.

Go deeper

PMLE Architecting Low-Code ML Solutions Practice Question

Why each option matters

Option-by-option breakdown

Related to this question