Q01 - Question
A shared services team receives a mixed stream of invoices and receipts in a single intake queue. The application should submit each document to one endpoint and still have the correct extraction model process it. Which two approaches meet this requirement? Each correct answer presents a complete solution.
Domain: Implement information extraction solutions (10–15%) Type: Multiple choice
- A. Use a composed model that groups the individual extraction models
- B. Use a custom classification model to identify the document type and route it to the matching extraction model
- C. Use the read model to extract text and rely on it to select the correct fields
- D. Train one custom template model on invoices only and submit all documents to it
A and B are correct.
Explanation: Composed models group individual extraction models so that one endpoint handles multiple document types, and a custom classifier can identify the document type and route it to the matching extraction model.
C is incorrect: The read model returns text and language or print classification, so it does not select document-specific fields or route documents.
D is incorrect: A single template model trained on invoices only cannot extract the receipt fields, and template models assume a consistent visual layout.
Q02 - Question
You are configuring fields in an Azure AI Search index. The search application must let users restrict results to a selected author, display counts of documents grouped by author, and order results by publication date. The author and date values are present in the index. How should you configure the index fields to support these requirements?
Domain: Implement information extraction solutions (10–15%) Type: Single choice
- A. Configure the author field as key and the date field as searchable.
- B. Configure both fields as retrievable only.
- C. Configure the author field as sortable and the date field as facetable.
- D. Configure the author field as filterable and facetable, and the date field as sortable.
D is correct.
Explanation: Index fields can be configured with specific attributes. Restricting results requires the filterable attribute, grouping counts requires the facetable attribute, and ordering requires the sortable attribute.
A is incorrect: The key attribute uniquely identifies a document, and the searchable attribute supports full-text matching; neither supports facet counts or ordering.
B is incorrect: The retrievable attribute only returns the value in results; it does not support filtering, faceting, or sorting by the search engine.
C is incorrect: The attributes are applied to the wrong fields. The author field requires filterable and facetable attributes, while the date field requires the sortable attribute.
Q03 - Question
A solution architect wants an analyst with no coding experience to label sample forms, train a custom model, review its accuracy, and test the model against a new document before developers integrate it. Which approach should the architect recommend?
Domain: Implement information extraction solutions (10–15%) Type: Single choice
- A. Create a custom project in Document Intelligence Studio
- B. Write a REST client that uploads labels and starts training
- C. Use the prebuilt invoice model and accept its default fields
- D. Build a composed model before any extraction models are trained
A is correct.
Explanation: Document Intelligence Studio provides a visual way to create a custom project that links the resource to a Blob Storage container, label sample forms, train the model, review accuracy, and test a new document.
B is incorrect: Writing a REST client requires coding, which the analyst cannot do.
C is incorrect: The prebuilt invoice model extracts its own fields without training, so it does not support labeling and accuracy review for custom forms.
D is incorrect: A composed model routes documents to extraction models, so the extraction models must exist first.
Q04 - Question
A business intelligence team needs to analyze entities and key phrases produced by an enrichment pipeline. They plan to use a reporting tool that requires a relational schema. The enrichment pipeline currently populates an Azure AI Search index. Which feature should you configure to support the reporting tool?
Domain: Implement information extraction solutions (10–15%) Type: Single choice
- A. A vector index that stores the enriched values.
- B. A knowledge store configured with a table projection.
- C. A knowledge store configured with a file projection.
- D. A knowledge store configured with a JSON object projection.
B is correct.
Explanation: A knowledge store persists projections from an enrichment pipeline. A table projection produces a relational schema, which meets the requirement for the reporting tool.
A is incorrect: A vector index supports vector-based RAG grounding and retrieval; it does not persist the enriched fields as relational tables for reporting tools.
C is incorrect: File projections persist image files extracted from documents, not the relational representation of enriched entities and key phrases.
D is incorrect: A JSON object projection persists data for ETL pipelines in a JSON format, not as a relational schema.
Q05 - Question
A development team plans to integrate Azure Document Intelligence into a claims processing application. The team needs the analysis results in a machine-readable format that includes the position of each extracted element on the page, so the application can highlight the source region for every value it stores. Which characteristic of the service output meets this requirement?
Domain: Implement information extraction solutions (10–15%) Type: Single choice
- A. Structured JSON that includes text, key-value pairs, tables, and bounding-box data
- B. A plain text transcript of the document with no positional information
- C. A rendered image file with the extracted values drawn onto the page
- D. A relational database table that the application queries directly
A is correct.
Explanation: Azure Document Intelligence returns structured JSON that contains text, key-value pairs, selection marks, tables, and bounding-box data, so your application can map each extracted value back to its position on the page.
B is incorrect: A plain text transcript omits the bounding-box data needed to highlight source regions.
C is incorrect: The output is structured JSON, not a rendered image with values drawn on the page.
D is incorrect: The service returns JSON through REST, SDKs, Studio, or Foundry, not a relational database table.
Q06 - Question
An operations team must process scanned compliance checklists that contain checkboxes, multi-column tables, and headings. The team wants to capture the table structure and the state of each checkbox without training a model. Which prebuilt option should the team use?
Domain: Implement information extraction solutions (10–15%) Type: Single choice
- A. The read model
- B. The layout model
- C. The identification document model
- D. A custom neural model
B is correct.
Explanation: The layout model extracts selection marks, tables, and document structure, fitting checklists that contain checkboxes and tables with no training required.
A is incorrect: The read model returns text and language or print classification, but it does not add selection marks or table structure.
C is incorrect: The identification document model targets identity documents, not compliance checklists.
D is incorrect: A custom neural model requires you to label sample forms and train the model, which does not meet the requirement to avoid training.
Q07 - Question
A finance department receives purchase requisitions from many suppliers. Each supplier uses a different page design, and field positions vary widely between documents. The department has capacity for a longer training cycle and needs the highest extraction quality across these varying designs. Which custom model type should the department train?
Domain: Implement information extraction solutions (10–15%) Type: Single choice
- A. A custom template model
- B. A custom classification model
- C. A custom neural model
- D. The prebuilt read model
C is correct.
Explanation: Custom neural models provide stronger results for documents with variable layouts, and the department can accept the additional training resources that neural training requires.
A is incorrect: Custom template models are intended for forms with a consistent visual layout.
B is incorrect: A classification model identifies the document type and must be paired with extraction models to return field values.
D is incorrect: The read model returns text and language or print classification, so it does not extract the labeled requisition fields.
Q08 - Question
A team uses Azure AI Search to index scanned agreements. The solution must extract text from the scanned page images, detect key phrases in that text, and extract custom structured fields that no built-in skill produces. Which two design actions should you include in the enrichment pipeline? Each correct answer presents part of the solution.
Domain: Implement information extraction solutions (10–15%) Type: Multiple choice
- A. Query the index using full Lucene syntax to extract text at query time.
- B. Configure built-in AI skills that use Foundry Tools capabilities for OCR and key phrase detection.
- C. Configure a table projection in a knowledge store to generate the missing structured fields.
- D. Configure a custom skill that wraps a service such as Document Intelligence to produce the missing structured fields.
B and D are correct.
Explanation: Built-in skills use Foundry Tools capabilities for tasks such as key-phrase detection and OCR. When no built-in skill produces the necessary fields, a custom skill can wrap services such as Document Intelligence to extract the custom structured fields.
A is incorrect: Queries evaluate an existing index through parsing and retrieval; they do not perform OCR extraction from images.
C is incorrect: A table projection persists enriched data in a relational schema for reporting; it does not generate new enriched fields during the pipeline process.
Q09 - Question
A company stores contracts as PDF files, product data in a relational database, and support notes as free text. The team requires one solution that can index all three content types, enrich the indexed content with AI, and supply grounding data for a retrieval augmented generation (RAG) application. Which capability of Azure AI Search supports using a single platform for this workload?
Domain: Implement information extraction solutions (10–15%) Type: Single choice
- A. Azure AI Search indexes structured, semi-structured, and unstructured data, enriches data with AI skills, and supports vector-based RAG grounding.
- B. Azure AI Search indexes structured and unstructured data natively but requires an external service to enrich the data with AI skills.
- C. Azure AI Search enriches unstructured data with AI skills but cannot index structured relational data.
- D. Azure AI Search supports enterprise search and knowledge mining but relies on an external service for vector-based RAG grounding.
A is correct.
Explanation: Azure AI Search indexes structured, semi-structured, and unstructured data, can enrich that data with AI skills, and supports enterprise search, knowledge mining, and vector-based grounding for RAG scenarios.
B is incorrect: Azure AI Search does not require an external service for enrichment; it uses an internal enrichment pipeline with AI skills.
C is incorrect: Azure AI Search supports indexing structured data in addition to semi-structured and unstructured data.
D is incorrect: Azure AI Search natively supports vector-based RAG grounding and does not rely on an external service for this capability.
Q10 - Question
You are designing an Azure AI Search solution. The solution must read documents from a data source, run an enrichment pipeline over the extracted content, and place both source fields and skill output fields into an index. Which component should you configure to perform these tasks?
Domain: Implement information extraction solutions (10–15%) Type: Single choice
- A. A knowledge store that writes projections to storage.
- B. A search query that uses full Lucene syntax.
- C. An indexer that connects to the data source.
- D. An OData filter expression.
C is correct.
Explanation: An Azure AI Search indexer connects to a data source, applies document cracking and an enrichment pipeline, and maps source and skill output fields to an index.
A is incorrect: A knowledge store persists projections of enriched data to storage; it does not read from the data source or populate the index.
B is incorrect: A search query evaluates search expressions against an existing index; it does not extract or enrich data.
D is incorrect: An OData filter expression evaluates criteria to filter query results; it is not used to extract or enrich source data.