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 10

Question 90

Different reporting tools duplicate a sensitive sales join and occasionally disagree on calculations. Which test result BEST demonstrates that the intended solution is working?

  1. A migration test catches incompatible column modification before production.
  2. Both tools return the same verified sales metric from the view.
  3. Tests resolve the intended procedure and handle null/invalid inputs.
  4. A measured rewrite returns identical values with lower CPU.

Explanation

Correct Answer

B:

Both tools return the same verified sales metric from the view. This is the decisive observable evidence after implementing create a governed view projecting the approved join and columns. It tests the underlying requirement: A view centralizes a query contract without storing a second copy by default.

Incorrect Answers

A:

A migration test catches incompatible column modification before production. However, that result would not establish the required outcome: both tools return the same verified sales metric from the view.

C:

Tests resolve the intended procedure and handle null/invalid inputs. However, that result would not establish the required outcome: both tools return the same verified sales metric from the view.

D:

A measured rewrite returns identical values with lower CPU. However, that result would not establish the required outcome: both tools return the same verified sales metric from the view.

Question 91

A developer exposes a write-capable MCP database tool to an agent with no approval boundary. What is the MOST relevant technical distinction in this situation?

  1. GraphQL relationship traversal can increase data exposure if access is too broad.
  2. Embedding static credentials in SQL code can expose model access and billing.
  3. Agents can turn natural-language requests into real database mutations if tools permit.
  4. Identity permission at one Azure service does not imply access to adjacent resources.

Explanation

Correct Answer

C:

Agents can turn natural-language requests into real database mutations if tools permit. This explains why the corrective action is to scope MCP tool actions and require human confirmation for consequential changes. Confirm the diagnosis when a test request cannot alter billing data without approval.

Incorrect Answers

A:

GraphQL relationship traversal can increase data exposure if access is too broad. The actual technical constraint is that agents can turn natural-language requests into real database mutations if tools permit.

B:

Embedding static credentials in SQL code can expose model access and billing. The actual technical constraint is that agents can turn natural-language requests into real database mutations if tools permit.

D:

Identity permission at one Azure service does not imply access to adjacent resources. The actual technical constraint is that agents can turn natural-language requests into real database mutations if tools permit.

Question 92

A container app's managed identity can log in but cannot read the required database view. Which ongoing control is MOST relevant after the issue is fixed?

  1. Scope identity permissions and review token failures.
  2. Review view definitions and permission chains.
  3. Expire elevation and log the migration transaction.
  4. Track principal creation across environments.

Explanation

Correct Answer

D:

Track principal creation across environments. This guards against recurrence because azure identity authentication and database authorization are distinct checks. The direct remediation is: Create or map the appropriate principal and grant minimum SQL permissions.

Incorrect Answers

A:

Scope identity permissions and review token failures. The current task instead requires this action: track principal creation across environments.

B:

Review view definitions and permission chains. The current task instead requires this action: track principal creation across environments.

C:

Expire elevation and log the migration transaction. The current task instead requires this action: track principal creation across environments.

Question 93

A customer-support assistant must classify multilingual complaints with a small latency budget. What is the MOST appropriate immediate corrective approach?

  1. Evaluate multilingual compact and large models on representative language and latency benchmarks.
  2. Choose and test a multimodal-capable external model with documented input support.
  3. Select language-appropriate content and model capabilities while retaining locale as metadata.
  4. Use heading-aware and language-aware chunk boundaries with source metadata.

Explanation

Correct Answer

A:

Evaluate multilingual compact and large models on representative language and latency benchmarks. This targets the primary issue because a larger model is not automatically the best fit for cost, speed or multilingual accuracy. Verification: Blind tests meet per-language quality and p95 response targets. Prevent recurrence: Rebenchmark when model versions or language distribution change.

Incorrect Answers

B:

Text-only models cannot directly interpret image content without a supporting vision pipeline.

C:

Embedding the wrong language source can degrade multilingual retrieval quality.

D:

Using heading-aware and language-aware chunk boundaries with source metadata would not correct a customer-support assistant must classify multilingual complaints with a small latency budget. Evaluate multilingual compact and large models on representative language and latency benchmarks.

Question 94

A trigger on SalesLine inserts assumes only one row appears in inserted and updates one summary incorrectly. What observed evidence would BEST confirm successful remediation?

  1. Checkout commits within SLO while notifications eventually deliver.
  2. A 50-row INSERT updates all required summary totals correctly.
  3. The function composes correctly in a FROM clause.
  4. Injected failure rolls back all intended writes and preserves integrity.

Explanation

Correct Answer

B:

A 50-row INSERT updates all required summary totals correctly. This is the decisive observable evidence after implementing rewrite the DML trigger to handle the inserted/deleted pseudo-tables as sets. It tests the underlying requirement: SQL triggers run per statement and inserted can contain multiple rows.

Incorrect Answers

A:

Checkout commits within SLO while notifications eventually deliver. However, that result would not establish the required outcome: a 50-row INSERT updates all required summary totals correctly.

C:

The function composes correctly in a FROM clause. However, that result would not establish the required outcome: a 50-row INSERT updates all required summary totals correctly.

D:

Injected failure rolls back all intended writes and preserves integrity. However, that result would not establish the required outcome: a 50-row INSERT updates all required summary totals correctly.

Question 95

An app selects an enormous model for every short product tag and then exceeds budget. What is the MOST relevant technical distinction in this situation?

  1. Unconstrained prose is fragile to parse and can omit required keys.
  2. Model invocation can export inputs and incur charges, so object privileges matter.
  3. Model size increases cost and latency without guaranteeing proportional quality gains.
  4. External endpoint configuration can become stale independently of database code.

Explanation

Correct Answer

C:

Model size increases cost and latency without guaranteeing proportional quality gains. This explains why the corrective action is to benchmark smaller models and route complex cases only when quality needs justify it. Confirm the diagnosis when the tiered approach meets held-out accuracy with lower cost per accepted result.

Incorrect Answers

A:

Unconstrained prose is fragile to parse and can omit required keys. The actual technical constraint is that model size increases cost and latency without guaranteeing proportional quality gains.

B:

Modeling invocation can export inputs and incur charges, so object privileges matter would not correct an app selects an enormous model for every short product tag and then exceeds budget. Model size increases cost and latency without guaranteeing proportional quality gains.

D:

External endpoint configuration can become stale independently of database code. The actual technical constraint is that model size increases cost and latency without guaranteeing proportional quality gains.

Question 96

A billing schema is owned by finance but anyone in the repository can approve SQL changes. Which ongoing control is MOST relevant after the issue is fixed?

  1. Audit pipeline changes and approval records.
  2. Require branch protection and code review for production.
  3. Require reviews for schema changes.
  4. Keep ownership mapping updated when teams change.

Explanation

Correct Answer

D:

Keep ownership mapping updated when teams change. This guards against recurrence because path ownership helps route high-impact DDL for accountable review. The direct remediation is: Configure CODEOWNERS and required reviews for the finance database paths.

Incorrect Answers

A:

Audit pipeline changes and approval records. The current task instead requires this action: keep ownership mapping updated when teams change.

B:

Require branch protection and code review for production. The current task instead requires this action: keep ownership mapping updated when teams change.

C:

Requiring reviews for schema changes addresses a different database behavior; the relevant distinction is that keep ownership mapping updated when teams change.

Question 97

A junior engineer accepts an AI-generated DROP TABLE migration without checking dependencies. What is the MOST appropriate immediate corrective approach?

  1. Review the generated diff, run it in an isolated test database, and require approval.
  2. Remove and rotate exposed secrets; use approved enterprise AI surfaces and sanitized prompts.
  3. Connect the approved SQL MCP endpoint using supported secure transport and scoped credentials.
  4. Switch to an organization-approved identity and governed Copilot configuration.

Explanation

Correct Answer

A:

Review the generated diff, run it in an isolated test database, and require approval. This targets the primary issue because AI suggestions are proposals, not change-control authorization. Verification: Migration rehearsal retains required tables and supports rollback. Prevent recurrence: Do not grant production DDL permissions directly to AI tools.

Incorrect Answers

B:

AI chat providers and extensions may retain shared content outside the database boundary.

C:

An MCP client needs a discoverable trusted server and permissions to run tools.

D:

Personal and enterprise assistants can have different retention and policy controls.

Question 98

An Azure SQL procedure needs to send a controlled JSON prompt to an approved external model API. Which measurable outcome BEST verifies that the requirement is satisfied?

  1. The approved identity receives a 2xx response while an unauthorized one fails.
  2. A staging request returns the expected status and schema-valid response.
  3. A throttled test eventually succeeds or leaves a recoverable queue record.
  4. A cross-client negative test yields no unauthorized quoted text.

Explanation

Correct Answer

B:

A staging request returns the expected status and schema-valid response. This is the decisive observable evidence after implementing build the request payload and call sp_invoke_external_rest_endpoint through a supported secure configuration. It tests the underlying requirement: The stored procedure enables outbound HTTP but requires correct permissions, endpoint reachability and payload formatting.

Incorrect Answers

A:

The approved identity receives a 2xx response while an unauthorized one fails. However, that result would not establish the required outcome: a staging request returns the expected status and schema-valid response.

C:

A throttled test eventually succeeds or leaves a recoverable queue record. However, that result would not establish the required outcome: a staging request returns the expected status and schema-valid response.

D:

A cross-client negative test yields no unauthorized quoted text. However, that result would not establish the required outcome: a staging request returns the expected status and schema-valid response.

Question 99

A supplier feed sometimes sends malformed JSON in an otherwise valid text column. Which technical reason BEST explains the proposed correction?

  1. Parsing nested arrays into typed rows reduces fragile manual string extraction.
  2. Column-oriented compression and batch processing can accelerate broad aggregations.
  3. Malformed payloads can break downstream JSON_VALUE and OPENJSON processing.
  4. A surrogate primary key alone does not protect a separate business identifier.

Explanation

Correct Answer

C:

Malformed payloads can break downstream JSON_VALUE and OPENJSON processing. This explains why the corrective action is to validate JSON structure on ingestion using supported native JSON type or ISJSON checks where applicable. Confirm the diagnosis when invalid payloads are rejected into an error queue without corrupting accepted rows.

Incorrect Answers

A:

Parsing nested arrays into typed rows reduces fragile manual string extraction. The actual technical constraint is that malformed payloads can break downstream JSON_VALUE and OPENJSON processing.

B:

Column-oriented compression and batch processing can accelerate broad aggregations. The actual technical constraint is that malformed payloads can break downstream JSON_VALUE and OPENJSON processing.

D:

A surrogate primary key alone does not protect a separate business identifier. The actual technical constraint is that malformed payloads can break downstream JSON_VALUE and OPENJSON processing.

Question 100

A publicly exposed DAB instance floods SQL with unauthenticated expensive read requests. What maintenance or governance control BEST reduces recurrence of this problem?

  1. Monitor connection failures and restrict ingress.
  2. Block accidental production connection strings in repository review.
  3. Use consistent ordering and monitor costly filters.
  4. Use telemetry and abuse alerting across API and SQL.

Explanation

Correct Answer

D:

Use telemetry and abuse alerting across API and SQL. This guards against recurrence because autoscaling the API does not eliminate downstream database exhaustion or unauthorized access. The direct remediation is: Harden ingress, authentication, rate limiting and query bounds before production exposure.

Incorrect Answers

A:

Monitor connection failures and restrict ingress. The current task instead requires this action: use telemetry and abuse alerting across API and SQL.

B:

Block accidental production connection strings in repository review. The current task instead requires this action: use telemetry and abuse alerting across API and SQL.

C:

Use consistent ordering and monitor costly filters. The current task instead requires this action: use telemetry and abuse alerting across API and SQL.

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