AI-102 Implement generative AI solutions Practice Question
Your organization uses Azure AI Document Intelligence to extract data from invoices. The extraction accuracy for total amounts is low. You have a labeled dataset of 500 invoices. You need to improve the model's accuracy for the 'total amount' field. What should you do?
⚠ Common exam trap
Microsoft often tests the misconception that adjusting confidence thresholds or adding more predefined models can improve extraction accuracy, when in fact only custom training with labeled data addresses field-specific low accuracy.
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
✓
Create a custom neural model and train it with the labeled dataset.
Azure AI Document Intelligence's custom neural model is specifically designed to improve extraction accuracy for fields like 'total amount' by training on labeled datasets. Unlike the prebuilt invoice model, a custom neural model learns the unique layout and variations in your invoices, directly addressing low accuracy for a specific field. Training with 500 labeled invoices provides sufficient data to fine-tune the model's extraction capabilities.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Add additional predefined models for invoice processing.
Why it's wrong here
Adding further predefined invoice models supplies no new training signal for your invoice layout, so the total-amount field stays inaccurate. Predefined models suit quick, generic extraction when you lack labelled data; here 500 labelled invoices already exist for a custom model.
- ✗
Enable OCR enhancement to improve text recognition.
Why it's wrong here
OCR enhancement improves character recognition on noisy scans, but the invoices already yield readable text; the low accuracy on a specific field stems from insufficient labelled examples or model training, not recognition failure. It is tempting because OCR quality affects extraction, and it would be correct for poorly scanned or handwritten documents.
- ✗
Increase the confidence threshold for the total amount field.
Why it's wrong here
Raising the confidence threshold only suppresses low-confidence extractions; it cannot teach the model the correct total-amount pattern, so recall drops without accuracy gains. Threshold tuning suits filtering predictions in production, not remedying a model whose field extraction is genuinely weak.
- ✓
Create a custom neural model and train it with the labeled dataset.
Why this is correct
A custom neural model learns field-specific patterns from your labelled invoices, including layout and contextual cues around the total amount, which the prebuilt model handles poorly. Training it on the 500 labelled samples directly targets the low-accuracy field, satisfying the requirement to improve extraction for 'total amount'.
Go deeper
Related to this question
About these practice questions
One of 761 original AI-102 practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This AI-102 practice question is part of Courseiva's free Microsoft certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AI-102 exam.