McAfee Secure

Exam Code: DP-800

Exam Name: Developing AI-Enabled Database Solutions

Certification Provider: Microsoft

Corresponding Certification: Microsoft Certified: SQL AI Developer Associate

Microsoft DP-800 Questions & Answers

Study with Up-To-Date REAL Exam Questions and Answers from the ACTUAL Test

101 Questions & Answers with Testing Engine

"Developing AI-Enabled Database Solutions Exam", also known as DP-800 exam, is a Microsoft certification exam.

Pass your tests with the always up-to-date DP-800 Exam Engine. Your DP-800 training materials keep you at the head of the pack!

Satisfaction Guaranteed

Money Back Guarantee

Testking provides hassle-free money back guarantee with our products. That is because we have 100% trust in the abilities of our professional and experience product team, and our record is a proof of that.

99.6% PASS RATE
Was: $137.49
Now: $124.99

Microsoft DP-800 Practice Test 18

Question 170

Queries filter millions of product rows by a JSON property and repeatedly scan payloads. Which test result BEST demonstrates that the intended solution is working?

  1. A test payload creates the expected number of items with correct types.
  2. The execution plan uses the chosen path access strategy and returns matching products.
  3. A test reserves the requested count and avoids collisions.
  4. A maintenance rehearsal successfully switches a partition with required constraints.

Explanation

Correct Answer

B:

The execution plan uses the chosen path access strategy and returns matching products. This is the decisive observable evidence after implementing evaluate supported JSON index or indexed computed-column path for that SQL platform and version. It tests the underlying requirement: Unindexed JSON path predicates can incur expensive row-by-row parsing.

Incorrect Answers

A:

A test payload creates the expected number of items with correct types. However, that result would not establish the required outcome: the execution plan uses the chosen path access strategy and returns matching products.

C:

A test reserves the requested count and avoids collisions. However, that result would not establish the required outcome: the execution plan uses the chosen path access strategy and returns matching products.

D:

A maintenance rehearsal successfully switches a partition with required constraints. However, that result would not establish the required outcome: the execution plan uses the chosen path access strategy and returns matching products.

Question 171

An app embeds user queries with one model but documents with another and similarity search becomes erratic. Which technical principle BEST explains the relevant design decision?

  1. External model quotas and transient errors require controlled retry behavior.
  2. A database model definition binds SQL AI operations to a specific external provider configuration.
  3. Vectors from different embedding spaces are not reliably comparable by distance.
  4. Embedding the wrong language source can degrade multilingual retrieval quality.

Explanation

Correct Answer

C:

Vectors from different embedding spaces are not reliably comparable by distance. This explains why the corrective action is to use compatible embedding model/version for both corpus and query vectors. Confirm the diagnosis when a gold query set ranks known relevant chunks consistently after alignment.

Incorrect Answers

A:

External model quotas and transient errors require controlled retry behavior. The actual technical constraint is that vectors from different embedding spaces are not reliably comparable by distance.

B:

A database model definition binds SQL AI operations to a specific external provider configuration. The actual technical constraint is that vectors from different embedding spaces are not reliably comparable by distance.

D:

Embedding the wrong language source can degrade multilingual retrieval quality. The actual technical constraint is that vectors from different embedding spaces are not reliably comparable by distance.

Question 172

A fraud team seeks a two-hop relationship among accounts, devices and shared addresses. Following remediation, what safeguard BEST prevents a repeat of this specific failure?

  1. Guard against cyclic paths and unsupported graph features.
  2. Reject malformed payloads and account for missing paths.
  3. Validate format differences across target geographies.
  4. Test graph direction against curated fixtures.

Explanation

Correct Answer

D:

Test graph direction against curated fixtures. This guards against recurrence because reversing graph edge direction can change which relationships match. The direct remediation is: Use MATCH with explicit edge direction and aliases for the required path.

Incorrect Answers

A:

Guard against cyclic paths and unsupported graph features. The current task instead requires this action: test graph direction against curated fixtures.

B:

Reject malformed payloads and account for missing paths. The current task instead requires this action: test graph direction against curated fixtures.

C:

Validate format differences across target geographies. The current task instead requires this action: test graph direction against curated fixtures.

Question 173

A legal assistant must answer from a restricted set of client-specific contract clauses. Which technical action BEST resolves the primary problem?

  1. Use tenant-trimmed retrieval with citations and a strict abstention policy.
  2. Use deterministic SQL aggregation for arithmetic and add RAG only for explanatory source text if needed.
  3. Implement bounded retry/backoff with idempotency and dead-letter handling.
  4. Reconcile generated claims with SQL source values before publishing results.

Explanation

Correct Answer

A:

Use tenant-trimmed retrieval with citations and a strict abstention policy. This targets the primary issue because RAG is appropriate only if the retriever enforces contractual permissions. Verification: A cross-client negative test yields no unauthorized quoted text. Prevent recurrence: Audit access metadata and citation validity.

Incorrect Answers

B:

Numeric reconciliation is a structured calculation, not a reason to hallucinate an answer from documents.

C:

Rate limiting is often transient and should not erase durable database work.

D:

Reconciling generated claims with SQL source values before publishing results addresses a different database behavior; the relevant distinction is that use tenant-trimmed retrieval with citations and a strict abstention policy.

Question 174

A dashboard needs all line item identifiers aggregated into one JSON array per order. Which test result BEST demonstrates that the intended solution is working?

  1. Rows matching the expected category are returned in a controlled test.
  2. An order with three lines produces the correct three-element array.
  3. Queries find related entities after a validated data migration.
  4. Suppliers with no invoice are returned correctly with null-containing test data.

Explanation

Correct Answer

B:

An order with three lines produces the correct three-element array. This is the decisive observable evidence after implementing use JSON_ARRAYAGG grouped by the order identifier where supported. It tests the underlying requirement: JSON_ARRAYAGG can collect multiple row values into a JSON array.

Incorrect Answers

A:

Rows matching the expected category are returned in a controlled test. However, that result would not establish the required outcome: an order with three lines produces the correct three-element array.

C:

Queries find related entities after a validated data migration. However, that result would not establish the required outcome: an order with three lines produces the correct three-element array.

D:

Suppliers with no invoice are returned correctly with null-containing test data. However, that result would not establish the required outcome: an order with three lines produces the correct three-element array.

Question 175

A stored procedure integrates with a DAB endpoint but unit tests pass while the API returns wrong roles. What is the MOST relevant technical distinction in this situation?

  1. Schema compilation alone does not prove business rules are correct.
  2. Repository and job logs can leak plaintext secrets.
  3. Cross-system contracts can fail even when individual SQL procedures are correct.
  4. SDK-style builds depend on resolved tooling and compatible target platforms.

Explanation

Correct Answer

C:

Cross-system contracts can fail even when individual SQL procedures are correct. This explains why the corrective action is to add integration tests across database permissions, DAB configuration and HTTP behavior. Confirm the diagnosis when authorized and denied endpoint tests match the intended policy.

Incorrect Answers

A:

Schema compilation alone does not prove business rules are correct. The actual technical constraint is that cross-system contracts can fail even when individual SQL procedures are correct.

B:

Repository and job logs can leak plaintext secrets. The actual technical constraint is that cross-system contracts can fail even when individual SQL procedures are correct.

D:

SDK-style builds depend on resolved tooling and compatible target platforms. The actual technical constraint is that cross-system contracts can fail even when individual SQL procedures are correct.

Question 176

A developer enables broad verbose SQL tracing and production throughput falls sharply. Which ongoing control is MOST relevant after the issue is fixed?

  1. Retain logs immutably where required and limit audit administrators.
  2. Track principal creation across environments.
  3. Rehearse key recovery and avoid distributing keys in scripts.
  4. Review event scope and sensitive-data redaction periodically.

Explanation

Correct Answer

D:

Review event scope and sensitive-data redaction periodically. This guards against recurrence because high-volume blanket logging can impact throughput and reveal sensitive inputs. The direct remediation is: Use targeted audit groups and evaluate performance and retention costs.

Incorrect Answers

A:

Retain logs immutably where required and limit audit administrators. The current task instead requires this action: review event scope and sensitive-data redaction periodically.

B:

Tracking principal creation across environments addresses a different database behavior; the relevant distinction is that review event scope and sensitive-data redaction periodically.

C:

Rehearse key recovery and avoid distributing keys in scripts. The current task instead requires this action: review event scope and sensitive-data redaction periodically.

Question 177

An Azure-hosted worker stores a SQL username and password in its deployment manifest. What is the MOST appropriate immediate corrective approach?

  1. Use Microsoft Entra-based managed identity authentication where supported.
  2. Create or map the appropriate principal and grant minimum SQL permissions.
  3. Send audit output to appropriately protected centralized storage with restricted permissions.
  4. Scope MCP tool actions and require human confirmation for consequential changes.

Explanation

Correct Answer

A:

Use Microsoft Entra-based managed identity authentication where supported. This targets the primary issue because managed identities remove a static SQL password from the deployed application configuration. Verification: The worker connects with an approved identity and no stored SQL password. Prevent recurrence: Scope identity permissions and review token failures.

Incorrect Answers

B:

This action addresses a different case: A container app's managed identity can log in but cannot read the required database view. For that case, azure identity authentication and database authorization are distinct checks. For this case, it does not directly correct the observed problem; the distinguishing issue is managed identities remove a static SQL password from the deployed application configuration. The specifically relevant immediate technical action is: Use Microsoft Entra-based managed identity authentication where supported.

C:

Tamper-resistant off-host retention is stronger evidence than only local logs.

D:

Agents can turn natural-language requests into real database mutations if tools permit.

Question 178

A directory matches transposed contact first names such as 'Jonh' versus 'John'. What observed evidence would BEST confirm successful remediation?

  1. Known typo pairs rank near each other without automatically merging records.
  2. Candidate rankings improve for the known transpositions without false merges.
  3. Month-over-month differences match a reference table.
  4. Repeated execution assigns stable row numbers to equal-time events.

Explanation

Correct Answer

B:

Candidate rankings improve for the known transpositions without false merges. This is the decisive observable evidence after implementing test JARO_WINKLER_DISTANCE or related supported fuzzy metric against a reference dataset. It tests the underlying requirement: Jaro-Winkler-style measures can treat transpositions and prefixes differently from raw edits.

Incorrect Answers

A:

Known typo pairs rank near each other without automatically merging records. However, that result would not establish the required outcome: candidate rankings improve for the known transpositions without false merges.

C:

Month-over-month differences match a reference table. However, that result would not establish the required outcome: candidate rankings improve for the known transpositions without false merges.

D:

Repeated execution assigns stable row numbers to equal-time events. However, that result would not establish the required outcome: candidate rankings improve for the known transpositions without false merges.

Question 179

A system splits legal clauses every fixed 50 characters, separating negation words from conditions. What is the MOST relevant technical distinction in this situation?

  1. Oversized chunks mix unrelated topics and degrade passage-level relevance.
  2. Frequent numeric fields add update cost and can weaken semantic relevance.
  3. Splitting within essential context can reverse meaning and reduce retrieval accuracy.
  4. A larger model is not automatically the best fit for cost, speed or multilingual accuracy.

Explanation

Correct Answer

C:

Splitting within essential context can reverse meaning and reduce retrieval accuracy. This explains why the corrective action is to use structure-aware chunking and sensible overlap around boundaries. Confirm the diagnosis when a sampled retrieval retains the complete exclusion condition.

Incorrect Answers

A:

Oversized chunks mix unrelated topics and degrade passage-level relevance. The actual technical constraint is that splitting within essential context can reverse meaning and reduce retrieval accuracy.

B:

Frequent numeric fields add update cost and can weaken semantic relevance. The actual technical constraint is that splitting within essential context can reverse meaning and reduce retrieval accuracy.

D:

A larger model is not automatically the best fit for cost, speed or multilingual accuracy. The actual technical constraint is that splitting within essential context can reverse meaning and reduce retrieval accuracy.

Question 180

A SQL AI integration needs a managed definition for an approved text embedding endpoint. Which ongoing control is MOST relevant after the issue is fixed?

  1. Inspect invocation logs and rate limits.
  2. Check supported modalities, retention and safety policy.
  3. Alert on per-document cost and failed queue age.
  4. Version endpoint changes and restrict model creation rights.

Explanation

Correct Answer

D:

Version endpoint changes and restrict model creation rights. This guards against recurrence because a database model definition binds SQL AI operations to a specific external provider configuration. The direct remediation is: Create an external model object with supported endpoint, model identifiers and secure credential strategy.

Incorrect Answers

A:

Inspect invocation logs and rate limits. The current task instead requires this action: version endpoint changes and restrict model creation rights.

B:

Check supported modalities, retention and safety policy. The current task instead requires this action: version endpoint changes and restrict model creation rights.

C:

Alert on per-document cost and failed queue age. The current task instead requires this action: version endpoint changes and restrict model creation rights.

Developing AI-Enabled Database Solutions (DP-800) Practice Tests