Microsoft AI-102 (Designing and Implementing a Microsoft Azure AI Solution) Practice Test 49

Question 480

Greeting questions start displacing more specific FAQ answers. What should be inspected?

  1. Group equivalent phrasings under one authoritative answer instead of cloning content entries
  2. Check the follow-up transition and answer selection conditions for that branch
  3. Version the approved FAQ content and the deployment promoted to production
  4. Review chit-chat matching behavior and remove overly broad conversational triggers

Explanation

Correct Answer

D:

Over-generalized greeting entries can compete with real help requests. In the employee FAQ chat interface, confirm the fix by retesting the affected behavior after reviewing chit-chat matching behavior and remove overly broad conversational triggers.

Incorrect Answers

A:

Duplicate answer records increase maintenance risk; one answer with alternates keeps updates consistent in its own workflow, but that is a different correction from the present requirement to review chit-chat matching behavior and remove overly broad conversational triggers. The reported issue would remain.

B:

Although useful elsewhere in the employee FAQ chat interface, dialogue paths need a terminal or next-answer transition when their missing parameter is satisfied; the present case is different and calls for the team to review chit-chat matching behavior and remove overly broad conversational triggers.

C:

That change addresses another failure mode: content and deployment provenance establish which answer set was exposed at a given time; it does not address this case, for which the relevant intervention is to review chit-chat matching behavior and remove overly broad conversational triggers.

Question 481

In an enterprise policy-document search service, engineers must describe the supported external store from which an indexer will ingest content. The expected evidence is that the indexer has a defined source collection from which to read. Which implementation is the best match?

  1. Configure an indexer data source
  2. Define an enrichment skillset
  3. Define an Azure AI Search index schema
  4. Configure knowledge store projections

Explanation

Correct Answer

A:

A data source supplies a connection and source configuration for supported Azure AI Search indexing pipelines. For this enterprise policy-document search service, the intended evidence is that the indexer has a defined source collection from which to read. It does not define the searchable field mapping for an index by itself.

Incorrect Answers

B:

A skillset chains built-in or custom AI skills to produce enriched fields from source content; it may be used to run OCR, entity extraction, or other supported enrichment steps during indexing in a enterprise policy-document search service, but the current requirement is instead to describe the supported external store from which an indexer will ingest content. That separate capability does not prove that the indexer has a defined source collection from which to read; A skillset does not replace the target index or the indexer's source connection, so selecting that alternative would not meet the stated acceptance check.

C:

An index schema defines field types and attributes controlling indexing and query behavior; one suitable field serves as the document key; it may be used to establish searchable, filterable, sortable, and key fields for the queryable document collection in a enterprise policy-document search service, but the current requirement is instead to describe the supported external store from which an indexer will ingest content. That separate capability does not prove that the indexer has a defined source collection from which to read; A data-source connection alone does not specify which fields are filterable or searchable, so selecting that alternative would not meet the stated acceptance check.

D:

Knowledge store projections persist selected AI-enrichment outputs outside the search index in supported storage structures; it may be used to retain enrichment outputs as projected objects, files, or tables for downstream analysis in a enterprise policy-document search service, but the current requirement is instead to describe the supported external store from which an indexer will ingest content. That separate capability does not prove that the indexer has a defined source collection from which to read; They are not equivalent to retrieving ranked documents through the search endpoint, so selecting that alternative would not meet the stated acceptance check.

Question 482

A data source points at a similarly named container in the wrong account. What should be corrected?

  1. Define the intended knowledge store objects or table projections and their destination
  2. Configure output field mappings or projections from enrichment nodes to index fields
  3. Verify the connection string or identity and source container reference
  4. Monitor indexer execution history and alert on failed runs

Explanation

Correct Answer

C:

A valid but incorrect source reference can ingest the wrong corpus without causing an obvious schema error. In the enterprise policy-document search service, confirm the fix by retesting the affected behavior after verifying the connection string or identity and source container reference.

Incorrect Answers

A:

Enrichment applied to an index does not automatically create external reporting artifacts in its own workflow, but that is a different correction from the present requirement to verify the connection string or identity and source container reference. The reported issue would remain.

B:

That change addresses another failure mode: enrichment results are not searchable until mapped into stored index content; it does not address this case, for which the relevant intervention is to verify the connection string or identity and source container reference.

D:

Although useful elsewhere in the enterprise policy-document search service, a schedule initiates jobs, but it does not guarantee they complete successfully; the present case is different and calls for the team to verify the connection string or identity and source container reference.

Question 483

A team responsible for an enterprise policy-document search service needs to retain enrichment outputs as projected objects, files, or tables for downstream analysis; the acceptance check is that downstream processes can access projected enriched artifacts. Which capability should be selected?

  1. Configure knowledge store projections
  2. Define an enrichment skillset
  3. Configure an indexer data source
  4. Define an Azure AI Search index schema

Explanation

Correct Answer

A:

Knowledge store projections persist selected AI-enrichment outputs outside the search index in supported storage structures. For this enterprise policy-document search service, the intended evidence is that downstream processes can access projected enriched artifacts. They are not equivalent to retrieving ranked documents through the search endpoint.

Incorrect Answers

B:

A skillset chains built-in or custom AI skills to produce enriched fields from source content; it may be used to run OCR, entity extraction, or other supported enrichment steps during indexing in a enterprise policy-document search service, but the current requirement is instead to retain enrichment outputs as projected objects, files, or tables for downstream analysis. That separate capability does not prove that downstream processes can access projected enriched artifacts; A skillset does not replace the target index or the indexer's source connection, so selecting that alternative would not meet the stated acceptance check.

C:

A data source supplies a connection and source configuration for supported Azure AI Search indexing pipelines; it may be used to describe the supported external store from which an indexer will ingest content in a enterprise policy-document search service, but the current requirement is instead to retain enrichment outputs as projected objects, files, or tables for downstream analysis. That separate capability does not prove that downstream processes can access projected enriched artifacts; It does not define the searchable field mapping for an index by itself, so selecting that alternative would not meet the stated acceptance check.

D:

An index schema defines field types and attributes controlling indexing and query behavior; one suitable field serves as the document key; it may be used to establish searchable, filterable, sortable, and key fields for the queryable document collection in a enterprise policy-document search service, but the current requirement is instead to retain enrichment outputs as projected objects, files, or tables for downstream analysis. That separate capability does not prove that downstream processes can access projected enriched artifacts; A data-source connection alone does not specify which fields are filterable or searchable, so selecting that alternative would not meet the stated acceptance check.

Question 484

A release for an enterprise policy-document search service is blocked until the application can establish searchable, filterable, sortable, and key fields for the queryable document collection. The tester will verify that queries can address fields with the intended schema attributes. What should be implemented?

  1. Define an enrichment skillset
  2. Configure an indexer data source
  3. Create and schedule an indexer
  4. Configure knowledge store projections
  5. Define an Azure AI Search index schema

Explanation

Correct Answer

E:

An index schema defines field types and attributes controlling indexing and query behavior; one suitable field serves as the document key. For this enterprise policy-document search service, the intended evidence is that queries can address fields with the intended schema attributes. A data-source connection alone does not specify which fields are filterable or searchable.

Incorrect Answers

A:

A skillset chains built-in or custom AI skills to produce enriched fields from source content; it may be used to run OCR, entity extraction, or other supported enrichment steps during indexing in a enterprise policy-document search service, but the current requirement is instead to establish searchable, filterable, sortable, and key fields for the queryable document collection. That separate capability does not prove that queries can address fields with the intended schema attributes; A skillset does not replace the target index or the indexer's source connection, so selecting that alternative would not meet the stated acceptance check.

B:

A data source supplies a connection and source configuration for supported Azure AI Search indexing pipelines; it may be used to describe the supported external store from which an indexer will ingest content in a enterprise policy-document search service, but the current requirement is instead to establish searchable, filterable, sortable, and key fields for the queryable document collection. That separate capability does not prove that queries can address fields with the intended schema attributes; It does not define the searchable field mapping for an index by itself, so selecting that alternative would not meet the stated acceptance check.

C:

An indexer pulls data from configured sources, applies configured mappings or skills, and writes searchable documents; it may be used to automate extraction and refresh of supported source data into an Azure AI Search index in a enterprise policy-document search service, but the current requirement is instead to establish searchable, filterable, sortable, and key fields for the queryable document collection. That separate capability does not prove that queries can address fields with the intended schema attributes; A query request is not an indexing schedule and cannot refresh stale source records, so selecting that alternative would not meet the stated acceptance check.

D:

Knowledge store projections persist selected AI-enrichment outputs outside the search index in supported storage structures; it may be used to retain enrichment outputs as projected objects, files, or tables for downstream analysis in a enterprise policy-document search service, but the current requirement is instead to establish searchable, filterable, sortable, and key fields for the queryable document collection. That separate capability does not prove that queries can address fields with the intended schema attributes; They are not equivalent to retrieving ranked documents through the search endpoint, so selecting that alternative would not meet the stated acceptance check.

Question 485

A new document was uploaded to storage but has not appeared in search. The indexer runs once per day. What explains the delay?

  1. Run the indexer on demand or adjust its supported refresh schedule
  2. Set the field's `filterable` capability in a compatible index schema before indexing
  3. Add an OCR enrichment skill and project its extracted text into a searchable field
  4. Grant the indexer identity the required blob data-plane read permission on the intended storage scope

Explanation

Correct Answer

A:

Indexer scheduling determines when source changes are processed; storage updates do not necessarily push immediately. In the enterprise policy-document search service, confirm the fix by retesting the affected behavior after running the indexer on demand or adjust its supported refresh schedule.

Incorrect Answers

B:

Although useful elsewhere in the enterprise policy-document search service, full-text searchable and filterable flags serve different query behaviors; filter expressions require filterable fields; the present case is different and calls for the team to run the indexer on demand or adjust its supported refresh schedule.

C:

Image-only pages need recognized text before lexical search can match their contents in its own workflow, but that is a different correction from the present requirement to run the indexer on demand or adjust its supported refresh schedule. The reported issue would remain.

D:

That change addresses another failure mode: a data-source object describes an ingestion connection but does not grant that connection authorization, so the configured managed identity must have suitable blob data-plane read rights at the source scope; it does not address this case, for which the relevant intervention is to run the indexer on demand or adjust its supported refresh schedule.

Question 486

A new index has no document key and ingestion fails to identify records uniquely. What must the schema include?

  1. Define the intended knowledge store objects or table projections and their destination
  2. Define a unique document key field of the supported key type
  3. Verify the connection string or identity and source container reference
  4. Configure output field mappings or projections from enrichment nodes to index fields

Explanation

Correct Answer

B:

Azure AI Search documents require an identifier to add, merge, or delete specific records. In the enterprise policy-document search service, confirm the fix by retesting the affected behavior after defining a unique document key field of the supported key type.

Incorrect Answers

A:

Enrichment applied to an index does not automatically create external reporting artifacts in its own workflow, but that is a different correction from the present requirement to define a unique document key field of the supported key type. The reported issue would remain.

C:

Although useful elsewhere in the enterprise policy-document search service, a valid but incorrect source reference can ingest the wrong corpus without causing an obvious schema error; the present case is different and calls for the team to define a unique document key field of the supported key type.

D:

That change addresses another failure mode: enrichment results are not searchable until mapped into stored index content; it does not address this case, for which the relevant intervention is to define a unique document key field of the supported key type.

Question 487

A reporting team needs OCR-enriched tables, not opaque JSON blobs, in its data lake. Which projection design consideration matters?

  1. Add an OCR enrichment skill and project its extracted text into a searchable field
  2. Run the indexer on demand or adjust its supported refresh schedule
  3. Set the field's `filterable` capability in a compatible index schema before indexing
  4. Choose a supported knowledge store projection shape matching downstream table analytics

Explanation

Correct Answer

D:

Projection formats influence how consumers query enriched content; selecting a table-oriented shape is different from merely enabling enrichment. In the enterprise policy-document search service, confirm the fix by retesting the affected behavior after choosing a supported knowledge store projection shape matching downstream table analytics.

Incorrect Answers

A:

That change addresses another failure mode: image-only pages need recognized text before lexical search can match their contents; it does not address this case, for which the relevant intervention is to choose a supported knowledge store projection shape matching downstream table analytics.

B:

Although useful elsewhere in the enterprise policy-document search service, indexer scheduling determines when source changes are processed; storage updates do not necessarily push immediately; the present case is different and calls for the team to choose a supported knowledge store projection shape matching downstream table analytics.

C:

Full-text searchable and filterable flags serve different query behaviors; filter expressions require filterable fields in its own workflow, but that is a different correction from the present requirement to choose a supported knowledge store projection shape matching downstream table analytics. The reported issue would remain.

Question 488

A skillset adds a custom model prediction that frequently times out during indexing. What should be investigated?

  1. Include stable source-document identifiers with stored enrichment outputs
  2. Use a supported managed identity and grant it the required storage data access
  3. Review custom skill endpoint latency, retry behavior, and indexing batch settings
  4. Configure supported deletion detection or an explicit index document deletion process

Explanation

Correct Answer

C:

Custom enrichment calls are external dependencies with resource limits and error-handling requirements. In the enterprise policy-document search service, confirm the fix by retesting the affected behavior after reviewing custom skill endpoint latency, retry behavior, and indexing batch settings.

Incorrect Answers

A:

Although useful elsewhere in the enterprise policy-document search service, provenance makes projected artifacts traceable and supports correction or deletion of originating content; the present case is different and calls for the team to review custom skill endpoint latency, retry behavior, and indexing batch settings.

B:

Having a data-source definition is not equivalent to possessing authorization on the external data store in its own workflow, but that is a different correction from the present requirement to review custom skill endpoint latency, retry behavior, and indexing batch settings. The reported issue would remain.

D:

That change addresses another failure mode: source removal does not automatically imply a corresponding index delete in every data-source configuration; it does not address this case, for which the relevant intervention is to review custom skill endpoint latency, retry behavior, and indexing batch settings.

Question 489

PDF scans index with no searchable text because the source pages are images. Which enrichment step is needed?

  1. Grant the indexer identity the required blob data-plane read permission on the intended storage scope
  2. Choose a supported knowledge store projection shape matching downstream table analytics
  3. Run the indexer on demand or adjust its supported refresh schedule
  4. Add an OCR enrichment skill and project its extracted text into a searchable field

Explanation

Correct Answer

D:

Image-only pages need recognized text before lexical search can match their contents. In the enterprise policy-document search service, confirm the fix by retesting the affected behavior after adding an OCR enrichment skill and project its extracted text into a searchable field.

Incorrect Answers

A:

A data-source object describes an ingestion connection but does not grant that connection authorization, so the configured managed identity must have suitable blob data-plane read rights at the source scope in its own workflow, but that is a different correction from the present requirement to add an OCR enrichment skill and project its extracted text into a searchable field. The reported issue would remain.

B:

Although useful elsewhere in the enterprise policy-document search service, projection formats influence how consumers query enriched content; selecting a table-oriented shape is different from merely enabling enrichment; the present case is different and calls for the team to add an OCR enrichment skill and project its extracted text into a searchable field.

C:

That change addresses another failure mode: indexer scheduling determines when source changes are processed; storage updates do not necessarily push immediately; it does not address this case, for which the relevant intervention is to add an OCR enrichment skill and project its extracted text into a searchable field.

Question 490

Deleted source files remain searchable after later indexing runs. What should be evaluated?

  1. Choose field attributes and query behavior appropriate for exact metadata versus analyzed text
  2. Review custom skill endpoint latency, retry behavior, and indexing batch settings
  3. Use a supported managed identity and grant it the required storage data access
  4. Configure supported deletion detection or an explicit index document deletion process

Explanation

Correct Answer

D:

Source removal does not automatically imply a corresponding index delete in every data-source configuration. In the enterprise policy-document search service, confirm the fix by retesting the affected behavior after configuring supported deletion detection or an explicit index document deletion process.

Incorrect Answers

A:

Identifiers and prose have different matching requirements; tokenization can change exact-match behavior in its own workflow, but that is a different correction from the present requirement to configure supported deletion detection or an explicit index document deletion process. The reported issue would remain.

B:

Although useful elsewhere in the enterprise policy-document search service, custom enrichment calls are external dependencies with resource limits and error-handling requirements; the present case is different and calls for the team to configure supported deletion detection or an explicit index document deletion process.

C:

That change addresses another failure mode: having a data-source definition is not equivalent to possessing authorization on the external data store; it does not address this case, for which the relevant intervention is to configure supported deletion detection or an explicit index document deletion process.

Designing and Implementing a Microsoft Azure AI Solution (AI-102) Practice Tests