AI-102 Question 391
Select 2You are tasked with building an Azure AI Document Intelligence solution for a financial company to process loan applications. The company receives scanned documents, including handwritten forms, printed forms, and invoices. They need to extract structured information such as names, addresses, loan amounts, and due dates. Which prebuilt Azure AI Document Intelligence model(s) would you use to meet this requirement?
- A
Prebuilt Receipt Model
- B
Prebuilt Invoice Model
- C
Prebuilt Read Model
- D
Prebuilt Business Card Model
- E
Custom Document Model
Show answer and explanation
Correct answers: B, C
Explanation
To meet the requirements, the Prebuilt Invoice Model should be used to process invoice documents, as it is specifically designed for this purpose. The Prebuilt Read Model is needed to handle text extraction from both handwritten and printed forms. Other prebuilt models or custom solutions are not applicable or necessary for this scenario.
- A. Incorrect.
The Prebuilt Receipt Model is specifically designed for extracting information from receipts, such as merchant names, transaction dates, and totals. It is not suitable for processing invoices, handwritten forms, or printed forms.
- B. Correct.
The Prebuilt Invoice Model is designed to extract structured data such as invoice numbers, dates, amounts, and vendor details. It is suitable for processing the invoices included in the loan applications.
- C. Correct.
The Prebuilt Read Model is capable of extracting both printed and handwritten text from scanned documents. It is essential for processing handwritten and printed forms in this scenario.
- D. Incorrect.
The Prebuilt Business Card Model is designed for extracting contact information such as names, job titles, emails, and phone numbers from business cards. It is not relevant to the loan application documents.
- E. Incorrect.
The Custom Document Model allows you to train a model tailored to specific document types. However, in this case, prebuilt models (Invoice Model and Read Model) already meet the requirements, so a custom model is unnecessary.