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 85

Question 840

Different reporting tools duplicate a sensitive sales join and occasionally disagree on calculations. Which technical action BEST resolves the primary problem?

  1. Review dependency effects and change the schema-bound view deliberately before modifying underlying objects.
  2. Use a stored procedure with an explicit transaction and appropriate TRY/CATCH rollback.
  3. Rewrite the DML trigger to handle the inserted/deleted pseudo-tables as sets.
  4. Create a governed view projecting the approved join and columns.

Explanation

Correct Answer

D:

Create a governed view projecting the approved join and columns. This targets the primary issue because a view centralizes a query contract without storing a second copy by default. Verification: Both tools return the same verified sales metric from the view. Prevent recurrence: Control grants on base tables and treat the view as an interface contract.

Incorrect Answers

A:

Reviewing dependency effects and change the schema-bound view deliberately before modifying underlying objects addresses a different database behavior; the relevant distinction is that create a governed view projecting the approved join and columns.

B:

Using a stored procedure with an explicit transaction and appropriate TRY/CATCH rollback addresses a different database behavior; the relevant distinction is that create a governed view projecting the approved join and columns.

C:

SQL triggers run per statement and inserted can contain multiple rows.

Question 841

A developer exposes a write-capable MCP database tool to an agent with no approval boundary. Which test result BEST demonstrates that the intended solution is working?

  1. A test request cannot alter billing data without approval.
  2. Cross-tenant nested queries return no unauthorized records in negative tests.
  3. A role-based test demonstrates restricted account values are masked.
  4. A spoofed request cannot change access to another tenant's rows.

Explanation

Correct Answer

A:

A test request cannot alter billing data without approval. This is the decisive observable evidence after implementing scope MCP tool actions and require human confirmation for consequential changes. It tests the underlying requirement: Agents can turn natural-language requests into real database mutations if tools permit.

Incorrect Answers

B:

Cross-tenant nested queries return no unauthorized records in negative tests. However, that result would not establish the required outcome: a test request cannot alter billing data without approval.

C:

A role-based test demonstrates restricted account values are masked. However, that result would not establish the required outcome: a test request cannot alter billing data without approval.

D:

A spoofed request cannot change access to another tenant's rows. However, that result would not establish the required outcome: a test request cannot alter billing data without approval.

Question 842

A container app's managed identity can log in but cannot read the required database view. Which technical principle BEST explains the relevant design decision?

  1. Managed identities remove a static SQL password from the deployed application configuration.
  2. Azure identity authentication and database authorization are distinct checks.
  3. Authorized query transformations and inference can expose sensitive values despite masking.
  4. DDM changes what unauthorized queries display; it is not data encryption.

Explanation

Correct Answer

B:

Azure identity authentication and database authorization are distinct checks. This explains why the corrective action is to create or map the appropriate principal and grant minimum SQL permissions. Confirm the diagnosis when the app reads the intended view but cannot read private financial objects.

Incorrect Answers

A:

Managed identities remove a static SQL password from the deployed application configuration. The actual technical constraint is that Azure identity authentication and database authorization are distinct checks.

C:

Authorized query transformations and inference can expose sensitive values despite masking. The actual technical constraint is that Azure identity authentication and database authorization are distinct checks.

D:

DDM changes what unauthorized queries display; it is not data encryption. The actual technical constraint is that Azure identity authentication and database authorization are distinct checks.

Question 843

A customer-support assistant must classify multilingual complaints with a small latency budget. Following remediation, what safeguard BEST prevents a repeat of this specific failure?

  1. Check supported modalities, retention and safety policy.
  2. Benchmark token limits and multilingual quality.
  3. Rebenchmark when model versions or language distribution change.
  4. Handle replay, deletion and watermark recovery.

Explanation

Correct Answer

C:

Rebenchmark when model versions or language distribution change. This guards against recurrence because a larger model is not automatically the best fit for cost, speed or multilingual accuracy. The direct remediation is: Evaluate multilingual compact and large models on representative language and latency benchmarks.

Incorrect Answers

A:

Check supported modalities, retention and safety policy. The current task instead requires this action: rebenchmark when model versions or language distribution change.

B:

Benchmark token limits and multilingual quality. The current task instead requires this action: rebenchmark when model versions or language distribution change.

D:

Handle replay, deletion and watermark recovery. The current task instead requires this action: rebenchmark when model versions or language distribution change.

Question 844

A trigger on SalesLine inserts assumes only one row appears in inserted and updates one summary incorrectly. Which technical action BEST resolves the primary problem?

  1. Move external notifications to a reliable asynchronous event workflow after durable commit.
  2. Consider converting eligible logic into an inline TVF and compare plans.
  3. Use parameterized stored procedures with scoped execution rights.
  4. Rewrite the DML trigger to handle the inserted/deleted pseudo-tables as sets.

Explanation

Correct Answer

D:

Rewrite the DML trigger to handle the inserted/deleted pseudo-tables as sets. This targets the primary issue because SQL triggers run per statement and inserted can contain multiple rows. Verification: A 50-row INSERT updates all required summary totals correctly. Prevent recurrence: Use multirow tests and avoid cursor-based per-row assumptions.

Incorrect Answers

A:

Synchronous trigger work participates in the calling transaction and can increase latency.

B:

Consider converting eligible logic into an inline TVF and compare plans. The relevant technical condition instead is that rewrite the DML trigger to handle the inserted/deleted pseudo-tables as sets.

C:

Using parameterized stored procedures with scoped execution rights addresses a different database behavior; the relevant distinction is that rewrite the DML trigger to handle the inserted/deleted pseudo-tables as sets.

Question 845

An app selects an enormous model for every short product tag and then exceeds budget. Which test result BEST demonstrates that the intended solution is working?

  1. The tiered approach meets held-out accuracy with lower cost per accepted result.
  2. Sample calls produce valid JSON schema instances or explicit errors.
  3. A gold query set ranks known relevant chunks consistently after alignment.
  4. An authorized test generates an embedding with the expected size.

Explanation

Correct Answer

A:

The tiered approach meets held-out accuracy with lower cost per accepted result. This is the decisive observable evidence after implementing benchmark smaller models and route complex cases only when quality needs justify it. It tests the underlying requirement: Model size increases cost and latency without guaranteeing proportional quality gains.

Incorrect Answers

B:

Sample calls produce valid JSON schema instances or explicit errors. However, that result would not establish the required outcome: the tiered approach meets held-out accuracy with lower cost per accepted result.

C:

A gold query set ranks known relevant chunks consistently after alignment. However, that result would not establish the required outcome: the tiered approach meets held-out accuracy with lower cost per accepted result.

D:

An authorized test generates an embedding with the expected size. However, that result would not establish the required outcome: the tiered approach meets held-out accuracy with lower cost per accepted result.

Question 846

A billing schema is owned by finance but anyone in the repository can approve SQL changes. Which technical principle BEST explains the relevant design decision?

  1. Automated execution without authorization can promote unreviewed schema changes.
  2. Path ownership helps route high-impact DDL for accountable review.
  3. SDK-style builds depend on resolved tooling and compatible target platforms.
  4. Repeatable deployments must not assume reference rows are absent.

Explanation

Correct Answer

B:

Path ownership helps route high-impact DDL for accountable review. This explains why the corrective action is to configure CODEOWNERS and required reviews for the finance database paths. Confirm the diagnosis when a billing SQL PR requires a finance-code-owner approval before merge.

Incorrect Answers

A:

Automated execution without authorization can promote unreviewed schema changes. The actual technical constraint is that path ownership helps route high-impact DDL for accountable review.

C:

SDK-style builds depend on resolved tooling and compatible target platforms. The actual technical constraint is that path ownership helps route high-impact DDL for accountable review.

D:

Repeatable deployments must not assume reference rows are absent. The actual technical constraint is that path ownership helps route high-impact DDL for accountable review.

Question 847

A junior engineer accepts an AI-generated DROP TABLE migration without checking dependencies. Which ongoing control is MOST relevant after the issue is fixed?

  1. Use pre-send secret scanning and developer training.
  2. Manage server allowlists and revoke unused tokens.
  3. Do not grant production DDL permissions directly to AI tools.
  4. Review grants and server capabilities after updates.

Explanation

Correct Answer

C:

Do not grant production DDL permissions directly to AI tools. This guards against recurrence because AI suggestions are proposals, not change-control authorization. The direct remediation is: Review the generated diff, run it in an isolated test database, and require approval.

Incorrect Answers

A:

Use pre-send secret scanning and developer training. The current task instead requires this action: do not grant production DDL permissions directly to AI tools.

B:

Manage server allowlists and revoke unused tokens. The current task instead requires this action: do not grant production DDL permissions directly to AI tools.

D:

Review grants and server capabilities after updates. The current task instead requires this action: do not grant production DDL permissions directly to AI tools.

Question 848

An Azure SQL procedure needs to send a controlled JSON prompt to an approved external model API. What is the MOST appropriate immediate corrective approach?

  1. Check REST endpoint authentication and configured identity before changing the SQL result parser.
  2. Parse the supported response schema and extract the correct generated content field.
  3. Serialize numeric SQL types as valid JSON numbers with explicit unit metadata.
  4. Build the request payload and call sp_invoke_external_rest_endpoint through a supported secure configuration.

Explanation

Correct Answer

D:

Build the request payload and call sp_invoke_external_rest_endpoint through a supported secure configuration. This targets the primary issue because the stored procedure enables outbound HTTP but requires correct permissions, endpoint reachability and payload formatting. Verification: A staging request returns the expected status and schema-valid response. Prevent recurrence: Use managed identity/secure credentials and enforce timeouts.

Incorrect Answers

A:

An authorization error occurs before model inference can produce valid content.

B:

HTTP response metadata and generated text are different parts of the payload.

C:

Serialize numeric SQL types as valid JSON numbers with explicit unit metadata. The relevant technical condition instead is that build the request payload and call sp_invoke_external_rest_endpoint through a supported secure configuration.

Question 849

A supplier feed sometimes sends malformed JSON in an otherwise valid text column. What observed evidence would BEST confirm successful remediation?

  1. Invalid payloads are rejected into an error queue without corrupting accepted rows.
  2. A test payload creates the expected number of items with correct types.
  3. A test reserves the requested count and avoids collisions.
  4. Plans show elimination for a month-filtered report and totals remain correct.

Explanation

Correct Answer

A:

Invalid payloads are rejected into an error queue without corrupting accepted rows. This is the decisive observable evidence after implementing validate JSON structure on ingestion using supported native JSON type or ISJSON checks where applicable. It tests the underlying requirement: Malformed payloads can break downstream JSON_VALUE and OPENJSON processing.

Incorrect Answers

B:

A test payload creates the expected number of items with correct types. However, that result would not establish the required outcome: invalid payloads are rejected into an error queue without corrupting accepted rows.

C:

A test reserves the requested count and avoids collisions. However, that result would not establish the required outcome: invalid payloads are rejected into an error queue without corrupting accepted rows.

D:

Plans show elimination for a month-filtered report and totals remain correct. However, that result would not establish the required outcome: invalid payloads are rejected into an error queue without corrupting accepted rows.

Question 850

A publicly exposed DAB instance floods SQL with unauthenticated expensive read requests. What is the MOST relevant technical distinction in this situation?

  1. Publishing the API does not grant the host network or SQL access automatically.
  2. Autoscaling the API does not eliminate downstream database exhaustion or unauthorized access.
  3. REST endpoint mapping supplies predictable URLs and HTTP operations.
  4. GraphQL can request related fields through a typed schema rather than repeated round trips.

Explanation

Correct Answer

B:

Autoscaling the API does not eliminate downstream database exhaustion or unauthorized access. This explains why the corrective action is to harden ingress, authentication, rate limiting and query bounds before production exposure. Confirm the diagnosis when load tests enforce limits and approved users retain service.

Incorrect Answers

A:

Publishing the API does not grant the host network or SQL access automatically. The actual technical constraint is that autoscaling the API does not eliminate downstream database exhaustion or unauthorized access.

C:

REST endpoint mapping supplies predictable URLs and HTTP operations. The actual technical constraint is that autoscaling the API does not eliminate downstream database exhaustion or unauthorized access.

D:

GraphQL can request related fields through a typed schema rather than repeated round trips. The actual technical constraint is that autoscaling the API does not eliminate downstream database exhaustion or unauthorized access.

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