Microsoft AB-100 Knowledge Prompt and Model Strategy Practice Test

 

Topic 03 covers Knowledge, Prompt and Model Strategy for Microsoft AB-100, using the current skills measured as of July 22, 2026 and primary Microsoft documentation. For broader exam preparation, review the AB-100 Exam Dumps page. These are original practice questions, and every option includes a scenario-specific explanation.

Question 1

A knowledge architecture team is reviewing knowledge, prompt and model strategy. The team has confirmed that knowledge changes frequently. It must now decide how to proceed. What is the best design choice?

1. Choose retrieval when knowledge changes frequently.
2. Choose external reach because users lack the intended Copilot surface is the decisive constraint.
3. Reject a small model because task breadth exceeds validation makes that approach unsuitable.
4. Choose a baseline comparison before investing in tuning.
5. Attach intended use and evaluation evidence to a library entry.

Correct Answer: A

 

Correct Answer

Answer A is correct because This is correct because it applies the reserved architecture decision—choose retrieval when knowledge changes frequently—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose external reach when users lack the intended copilot surface’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘reject a small model when task breadth exceeds validation’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose a baseline comparison before investing in tuning’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘attach intended use and evaluation evidence to a library entry’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 2

A production-readiness review of knowledge, prompt and model strategy gives the edge AI pilot new evidence. The team has confirmed that open-ended answers are unacceptable. It must now decide how to proceed. What is the best design choice?

1. Assess the maintenance cost of several specialized models.
2. Assess model ownership against maintenance capacity.
3. Evaluate extension fit for an existing enterprise assistant.
4. Assess a small model for a narrow task with measured quality.
5. Choose bounded knowledge when open-ended answers are unacceptable.

Correct Answer: E

 

Correct Answer

Answer E is correct because This is correct because it applies the reserved architecture decision—choose bounded knowledge when open-ended answers are unacceptable—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘assess the maintenance cost of several specialized models’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘assess model ownership against maintenance capacity’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘evaluate extension fit for an existing enterprise assistant’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘assess a small model for a narrow task with measured quality’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 3

A production-readiness review of knowledge, prompt and model strategy gives the edge AI pilot new evidence. Stakeholders are treating external facts and agent instructions as the same decision, which is obscuring the actual control boundary. What is the best design choice?

1. Version reusable prompts with accountable owners.
2. Treat instructions separately from untrusted business content and evaluate each with its own evidence.
3. Retain standard Copilot where customization adds no business value.
4. Separate external facts from agent instructions.
5. Evaluate resource constraints in an edge scenario.

Correct Answer: D

 

Correct Answer

Answer D is correct because This is correct because it applies the reserved architecture decision—separate external facts from agent instructions—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘version reusable prompts with accountable owners’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘separate instructions from untrusted business content’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘retain standard copilot where customization adds no business value’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘evaluate resource constraints in an edge scenario’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 4

A pilot for knowledge, prompt and model strategy reaches a decision point for the enterprise assistant program. The current solution leaves source precedence when enterprise policies conflict unresolved, and the gap is now affecting the stated business or technical requirement. Which approach best fits the requirement?

1. Prefer grounding over customization for changing company facts and validate the result against the stated requirement.
2. Compare specialized model latency against quality requirements.
3. Treat reusable templates separately from sensitive runtime values and evaluate each with its own evidence.
4. Compare the chosen approach with a baseline on representative prompts, sources, and maintenance effort before standardizing it.
5. Set source precedence when enterprise policies conflict.

Correct Answer: E

 

Correct Answer

Answer E is correct because This is correct because it applies the reserved architecture decision—set source precedence when enterprise policies conflict—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘prefer grounding over customization for changing company facts’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘compare specialized model latency against quality requirements’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘separate reusable templates from sensitive runtime values’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is a useful verification step after the primary architecture decision, but it does not by itself resolve the decision the scenario asks for. The design choice must be made first, and this evidence can then confirm that the chosen approach behaves as intended.

 

Question 5

Before a wider rollout of knowledge, prompt and model strategy, the model strategy group reviews the current evidence. The current solution leaves generation only where output variation is acceptable unresolved, and the gap is now affecting the stated business or technical requirement. What is the best design choice?

1. Reject a small model because task breadth exceeds validation makes that approach unsuitable.
2. Choose generation only where output variation is acceptable.
3. Consider a custom model for a persistent task-specific quality gap.
4. Define review gates for shared prompt publication.
5. Use examples to clarify an ambiguous task.

Correct Answer: B

 

Correct Answer

Answer B is correct because This is correct because it applies the reserved architecture decision—choose generation only where output variation is acceptable—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘reject a small model when task breadth exceeds validation’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘consider a custom model for a persistent task-specific quality gap’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘define review gates for shared prompt publication’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘use examples to clarify an ambiguous task’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 6

During a design workshop on knowledge, prompt and model strategy, the prompt standards board identifies a constraint. The review has identified a specific issue: extend existing Copilot when the workflow stays in Microsoft 365. Which action should the team take next?

1. Reject customization because representative training data is absent makes that approach unsuitable.
2. Assess the maintenance cost of several specialized models.
3. Set explicit uncertainty behavior when evidence is insufficient.
4. Extend existing Copilot when the workflow stays in Microsoft 365.
5. Retire redundant templates with conflicting behavior.

Correct Answer: D

 

Correct Answer

Answer D is correct because This is correct because it applies the reserved architecture decision—extend existing Copilot when the workflow stays in Microsoft 365—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘reject customization when representative training data is absent’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘assess the maintenance cost of several specialized models’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘set explicit uncertainty behavior when evidence is insufficient’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘retire redundant templates with conflicting behavior’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 7

A pilot for knowledge, prompt and model strategy reaches a decision point for the enterprise assistant program. The team has confirmed that orchestration must be controlled. It must now decide how to proceed. What should the architect recommend?

1. Choose a custom agent when orchestration must be controlled.
2. Limit context to information relevant to the decision.
3. Attach intended use and evaluation evidence to a library entry.
4. Treat instructions separately from untrusted business content and evaluate each with its own evidence.
5. Choose a baseline comparison before investing in tuning.

Correct Answer: A

 

Correct Answer

Answer A is correct because This is correct because it applies the reserved architecture decision—choose a custom agent when orchestration must be controlled—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘limit context to information relevant to the decision’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘attach intended use and evaluation evidence to a library entry’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘separate instructions from untrusted business content’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose a baseline comparison before investing in tuning’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 8

During a design workshop on knowledge, prompt and model strategy, the prompt standards board identifies a constraint. The team has confirmed that users lack the intended Copilot surface. It must now decide how to proceed. Which action should the team take next?

1. Assess a small model for a narrow task with measured quality.
2. Assess model ownership against maintenance capacity.
3. Choose external reach when users lack the intended Copilot surface.
4. Compare the chosen approach with a baseline on representative prompts, sources, and maintenance effort before standardizing it.
5. Specify required output structure for downstream consumption.

Correct Answer: C

 

Correct Answer

Answer C is correct because This is correct because it applies the reserved architecture decision—choose external reach when users lack the intended Copilot surface—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘assess a small model for a narrow task with measured quality’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘assess model ownership against maintenance capacity’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is a useful verification step after the primary architecture decision, but it does not by itself resolve the decision the scenario asks for. The design choice must be made first, and this evidence can then confirm that the chosen approach behaves as intended.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘specify required output structure for downstream consumption’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 9

A pilot for knowledge, prompt and model strategy reaches a decision point for the enterprise assistant program. Existing evidence is insufficient to make a safe release or architecture decision about extension fit for an existing enterprise assistant. Which decision is most appropriate?

1. Choose bounded knowledge because open-ended answers are unacceptable is the decisive constraint.
2. Version reusable prompts with accountable owners.
3. Use examples to clarify an ambiguous task.
4. Evaluate extension fit for an existing enterprise assistant.
5. Evaluate resource constraints in an edge scenario.

Correct Answer: D

 

Correct Answer

Answer D is correct because This is correct because it applies the reserved architecture decision—evaluate extension fit for an existing enterprise assistant—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose bounded knowledge when open-ended answers are unacceptable’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘version reusable prompts with accountable owners’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘use examples to clarify an ambiguous task’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘evaluate resource constraints in an edge scenario’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 10

A production-readiness review of knowledge, prompt and model strategy gives the edge AI pilot new evidence. The current solution leaves standard Copilot where customization adds no business value unresolved, and the gap is now affecting the stated business or technical requirement. Which decision is most appropriate?

1. Treat external facts separately from agent instructions and evaluate each with its own evidence.
2. Set explicit uncertainty behavior when evidence is insufficient.
3. Compare specialized model latency against quality requirements.
4. Retain standard Copilot where customization adds no business value.
5. Treat reusable templates separately from sensitive runtime values and evaluate each with its own evidence.

Correct Answer: D

 

Correct Answer

Answer D is correct because This is correct because it applies the reserved architecture decision—retain standard Copilot where customization adds no business value—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘separate external facts from agent instructions’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘set explicit uncertainty behavior when evidence is insufficient’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘compare specialized model latency against quality requirements’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘separate reusable templates from sensitive runtime values’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 11

A knowledge architecture team is reviewing knowledge, prompt and model strategy. The current solution leaves grounding over customization for changing company facts unresolved, and the gap is now affecting the stated business or technical requirement. Which action should the team take next?

1. Reject a small model because task breadth exceeds validation makes that approach unsuitable.
2. Limit context to information relevant to the decision.
3. Set source precedence when enterprise policies conflict.
4. Prefer grounding over customization for changing company facts.
5. Define review gates for shared prompt publication.

Correct Answer: D

 

Correct Answer

Answer D is correct because This is correct because it applies the reserved architecture decision—prefer grounding over customization for changing company facts—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘reject a small model when task breadth exceeds validation’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘limit context to information relevant to the decision’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘set source precedence when enterprise policies conflict’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘define review gates for shared prompt publication’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 12

The implementation is ready to proceed except for one unresolved boundary: a custom model for a persistent task-specific quality gap. What should the team do next?

1. Assess the maintenance cost of several specialized models.
2. Compare the chosen approach with a baseline on representative prompts, sources, and maintenance effort before standardizing it.
3. Retire redundant templates with conflicting behavior.
4. Choose retrieval because knowledge changes frequently is the decisive constraint.
5. Consider a custom model for a persistent task-specific quality gap.

Correct Answer: E

 

Correct Answer

Answer E is correct because This is correct because it applies the reserved architecture decision—consider a custom model for a persistent task-specific quality gap—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘assess the maintenance cost of several specialized models’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is a useful verification step after the primary architecture decision, but it does not by itself resolve the decision the scenario asks for. The design choice must be made first, and this evidence can then confirm that the chosen approach behaves as intended.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘retire redundant templates with conflicting behavior’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose retrieval when knowledge changes frequently’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 13

A pilot for knowledge, prompt and model strategy reaches a decision point for the enterprise assistant program. A proposed design would customization, but representative training data is absent. Which action should the team take next?

1. Choose bounded knowledge because open-ended answers are unacceptable is the decisive constraint.
2. Extend existing Copilot when the workflow stays in Microsoft 365.
3. Reject customization when representative training data is absent.
4. Treat instructions separately from untrusted business content and evaluate each with its own evidence.
5. Attach intended use and evaluation evidence to a library entry.

Correct Answer: C

 

Correct Answer

Answer C is correct because This is correct because it applies the reserved architecture decision—reject customization when representative training data is absent—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose bounded knowledge when open-ended answers are unacceptable’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘extend existing copilot when the workflow stays in microsoft 365’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘separate instructions from untrusted business content’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘attach intended use and evaluation evidence to a library entry’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 14

The architecture team must document a decision about a baseline comparison before investing in tuning. A prior pilot did not resolve this point. Which action is most appropriate?

1. Assess a small model for a narrow task with measured quality.
2. Specify required output structure for downstream consumption.
3. Choose a baseline comparison before investing in tuning.
4. Treat external facts separately from agent instructions and evaluate each with its own evidence.
5. Choose a custom agent because orchestration must be controlled is the decisive constraint.

Correct Answer: C

 

Correct Answer

Answer C is correct because This is correct because it applies the reserved architecture decision—choose a baseline comparison before investing in tuning—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘assess a small model for a narrow task with measured quality’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘specify required output structure for downstream consumption’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘separate external facts from agent instructions’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose a custom agent when orchestration must be controlled’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 15

Before a wider rollout of knowledge, prompt and model strategy, the model strategy group reviews the current evidence. Existing evidence is insufficient to make a safe release or architecture decision about model ownership against maintenance capacity. Which approach best fits the requirement?

1. Set source precedence when enterprise policies conflict.
2. Use examples to clarify an ambiguous task.
3. Choose external reach because users lack the intended Copilot surface is the decisive constraint.
4. Assess model ownership against maintenance capacity.
5. Evaluate resource constraints in an edge scenario.

Correct Answer: D

 

Correct Answer

Answer D is correct because This is correct because it applies the reserved architecture decision—assess model ownership against maintenance capacity—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘set source precedence when enterprise policies conflict’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘use examples to clarify an ambiguous task’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose external reach when users lack the intended copilot surface’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘evaluate resource constraints in an edge scenario’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 16

A knowledge architecture team is reviewing knowledge, prompt and model strategy. The unresolved requirement concerns accountable owners. The current design does not yet establish how reusable prompts should be handled. Which approach best fits the requirement?

1. Choose generation only where output variation is acceptable.
2. Compare the chosen approach with a baseline on representative prompts, sources, and maintenance effort before standardizing it.
3. Version reusable prompts with accountable owners.
4. Compare specialized model latency against quality requirements.
5. Set explicit uncertainty behavior when evidence is insufficient.

Correct Answer: C

 

Correct Answer

Answer C is correct because This is correct because it applies the reserved architecture decision—version reusable prompts with accountable owners—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose generation only where output variation is acceptable’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is a useful verification step after the primary architecture decision, but it does not by itself resolve the decision the scenario asks for. The design choice must be made first, and this evidence can then confirm that the chosen approach behaves as intended.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘compare specialized model latency against quality requirements’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘set explicit uncertainty behavior when evidence is insufficient’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 17

Before a wider rollout of knowledge, prompt and model strategy, the model strategy group reviews the current evidence. Stakeholders are treating reusable templates and sensitive runtime values as the same decision, which is obscuring the actual control boundary. What is the best design choice?

1. Reject a small model because task breadth exceeds validation makes that approach unsuitable.
2. Separate reusable templates from sensitive runtime values.
3. Extend existing Copilot when the workflow stays in Microsoft 365.
4. Limit context to information relevant to the decision.
5. Retain standard Copilot where customization adds no business value.

Correct Answer: B

 

Correct Answer

Answer B is correct because This is correct because it applies the reserved architecture decision—separate reusable templates from sensitive runtime values—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘reject a small model when task breadth exceeds validation’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘extend existing copilot when the workflow stays in microsoft 365’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘limit context to information relevant to the decision’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘retain standard copilot where customization adds no business value’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 18

The solution currently has no defensible decision for review gates for shared prompt publication. Which recommendation best closes that gap?

1. Choose retrieval because knowledge changes frequently is the decisive constraint.
2. Prefer grounding over customization for changing company facts and validate the result against the stated requirement.
3. Define review gates for shared prompt publication.
4. Choose a custom agent because orchestration must be controlled is the decisive constraint.
5. Assess the maintenance cost of several specialized models.

Correct Answer: C

 

Correct Answer

Answer C is correct because This is correct because it applies the reserved architecture decision—define review gates for shared prompt publication—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose retrieval when knowledge changes frequently’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘prefer grounding over customization for changing company facts’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose a custom agent when orchestration must be controlled’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘assess the maintenance cost of several specialized models’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 19

An earlier design assumed the wrong thing about redundant templates with conflicting behavior. Which decision should replace that assumption?

1. Choose bounded knowledge because open-ended answers are unacceptable is the decisive constraint.
2. Retire redundant templates with conflicting behavior.
3. Choose external reach because users lack the intended Copilot surface is the decisive constraint.
4. Treat instructions separately from untrusted business content and evaluate each with its own evidence.
5. Consider a custom model for a persistent task-specific quality gap.

Correct Answer: B

 

Correct Answer

Answer B is correct because This is correct because it applies the reserved architecture decision—retire redundant templates with conflicting behavior—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose bounded knowledge when open-ended answers are unacceptable’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose external reach when users lack the intended copilot surface’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘separate instructions from untrusted business content’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘consider a custom model for a persistent task-specific quality gap’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 20

A production-readiness review of knowledge, prompt and model strategy gives the edge AI pilot new evidence. The unresolved requirement concerns a library entry. The current design does not yet establish how intended use and evaluation evidence should be handled. Which action should the team take next?

1. Attach intended use and evaluation evidence to a library entry.
2. Compare the chosen approach with a baseline on representative prompts, sources, and maintenance effort before standardizing it.
3. Specify required output structure for downstream consumption.
4. Do not publish a reusable prompt, model, or knowledge pattern until its intended use, limits, and evaluation evidence are documented.
5. Treat external facts separately from agent instructions and evaluate each with its own evidence.

Correct Answer: A

 

Correct Answer

Answer A is correct because This is correct because it applies the reserved architecture decision—attach intended use and evaluation evidence to a library entry—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer B is incorrect because This is a useful verification step after the primary architecture decision, but it does not by itself resolve the decision the scenario asks for. The design choice must be made first, and this evidence can then confirm that the chosen approach behaves as intended.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘specify required output structure for downstream consumption’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This safeguard can be important in a broader production design, but it is a supporting control rather than the primary decision required by this scenario. Selecting it alone would leave the core architecture choice unresolved.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘separate external facts from agent instructions’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 21

A pilot for knowledge, prompt and model strategy reaches a decision point for the enterprise assistant program. Existing evidence is insufficient to make a safe release or architecture decision about a small model for a narrow task with measured quality. What should the architect recommend?

1. Retain standard Copilot where customization adds no business value.
2. Assess a small model for a narrow task with measured quality.
3. Choose a baseline comparison before investing in tuning.
4. Use examples to clarify an ambiguous task.
5. Set source precedence when enterprise policies conflict.

Correct Answer: B

 

Correct Answer

Answer B is correct because This is correct because it applies the reserved architecture decision—assess a small model for a narrow task with measured quality—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘retain standard copilot where customization adds no business value’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose a baseline comparison before investing in tuning’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘use examples to clarify an ambiguous task’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘set source precedence when enterprise policies conflict’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 22

The architecture council wants a testable decision for resource constraints in an edge scenario. Which option should it approve?

1. Set explicit uncertainty behavior when evidence is insufficient.
2. Assess model ownership against maintenance capacity.
3. Evaluate resource constraints in an edge scenario.
4. Choose generation only where output variation is acceptable.
5. Prefer grounding over customization for changing company facts and validate the result against the stated requirement.

Correct Answer: C

 

Correct Answer

Answer C is correct because This is correct because it applies the reserved architecture decision—evaluate resource constraints in an edge scenario—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘set explicit uncertainty behavior when evidence is insufficient’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘assess model ownership against maintenance capacity’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose generation only where output variation is acceptable’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘prefer grounding over customization for changing company facts’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 23

The solution currently has no defensible decision for specialized model latency against quality requirements. Which recommendation best closes that gap?

1. Version reusable prompts with accountable owners.
2. Extend existing Copilot when the workflow stays in Microsoft 365.
3. Limit context to information relevant to the decision.
4. Consider a custom model for a persistent task-specific quality gap.
5. Compare specialized model latency against quality requirements.

Correct Answer: E

 

Correct Answer

Answer E is correct because This is correct because it applies the reserved architecture decision—compare specialized model latency against quality requirements—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘version reusable prompts with accountable owners’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘extend existing copilot when the workflow stays in microsoft 365’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘limit context to information relevant to the decision’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘consider a custom model for a persistent task-specific quality gap’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 24

The AI platform working group is preparing an architecture decision record for knowledge, prompt and model strategy. A proposed design would a small model, but task breadth exceeds validation. What is the best design choice?

1. Choose retrieval because knowledge changes frequently is the decisive constraint.
2. Choose a custom agent because orchestration must be controlled is the decisive constraint.
3. Compare the chosen approach with a baseline on representative prompts, sources, and maintenance effort before standardizing it.
4. Reject a small model when task breadth exceeds validation.
5. Reject customization because representative training data is absent makes that approach unsuitable.

Correct Answer: D

 

Correct Answer

Answer D is correct because This is correct because it applies the reserved architecture decision—reject a small model when task breadth exceeds validation—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose retrieval when knowledge changes frequently’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose a custom agent when orchestration must be controlled’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is a useful verification step after the primary architecture decision, but it does not by itself resolve the decision the scenario asks for. The design choice must be made first, and this evidence can then confirm that the chosen approach behaves as intended.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘reject customization when representative training data is absent’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 25

A production-readiness review of knowledge, prompt and model strategy gives the edge AI pilot new evidence. Existing evidence is insufficient to make a safe release or architecture decision about the maintenance cost of several specialized models. What is the best design choice?

1. Choose bounded knowledge because open-ended answers are unacceptable is the decisive constraint.
2. Define review gates for shared prompt publication.
3. Choose external reach because users lack the intended Copilot surface is the decisive constraint.
4. Assess the maintenance cost of several specialized models.
5. Choose a baseline comparison before investing in tuning.

Correct Answer: D

 

Correct Answer

Answer D is correct because This is correct because it applies the reserved architecture decision—assess the maintenance cost of several specialized models—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose bounded knowledge when open-ended answers are unacceptable’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘define review gates for shared prompt publication’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose external reach when users lack the intended copilot surface’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose a baseline comparison before investing in tuning’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 26

Before a wider rollout of knowledge, prompt and model strategy, the model strategy group reviews the current evidence. Stakeholders are treating instructions and untrusted business content as the same decision, which is obscuring the actual control boundary. Which decision is most appropriate?

1. Separate instructions from untrusted business content.
2. Retire redundant templates with conflicting behavior.
3. Evaluate extension fit for an existing enterprise assistant.
4. Assess model ownership against maintenance capacity.
5. Treat external facts separately from agent instructions and evaluate each with its own evidence.

Correct Answer: A

 

Correct Answer

Answer A is correct because This is correct because it applies the reserved architecture decision—separate instructions from untrusted business content—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘retire redundant templates with conflicting behavior’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘evaluate extension fit for an existing enterprise assistant’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘assess model ownership against maintenance capacity’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘separate external facts from agent instructions’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 27

The AI platform working group is preparing an architecture decision record for knowledge, prompt and model strategy. The unresolved requirement concerns downstream consumption. The current design does not yet establish how required output structure should be handled. Which decision is most appropriate?

1. Specify required output structure for downstream consumption.
2. Set source precedence when enterprise policies conflict.
3. Retain standard Copilot where customization adds no business value.
4. Attach intended use and evaluation evidence to a library entry.
5. Version reusable prompts with accountable owners.

Correct Answer: A

 

Correct Answer

Answer A is correct because This is correct because it applies the reserved architecture decision—specify required output structure for downstream consumption—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘set source precedence when enterprise policies conflict’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘retain standard copilot where customization adds no business value’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘attach intended use and evaluation evidence to a library entry’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘version reusable prompts with accountable owners’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 28

A review board has narrowed the remaining issue to examples to clarify an ambiguous task. Which approach best addresses that issue without broadening scope?

1. Compare the chosen approach with a baseline on representative prompts, sources, and maintenance effort before standardizing it.
2. Prefer grounding over customization for changing company facts and validate the result against the stated requirement.
3. Use examples to clarify an ambiguous task.
4. Choose generation only where output variation is acceptable.
5. Treat reusable templates separately from sensitive runtime values and evaluate each with its own evidence.

Correct Answer: C

 

Correct Answer

Answer C is correct because This is correct because it applies the reserved architecture decision—use examples to clarify an ambiguous task—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is a useful verification step after the primary architecture decision, but it does not by itself resolve the decision the scenario asks for. The design choice must be made first, and this evidence can then confirm that the chosen approach behaves as intended.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘prefer grounding over customization for changing company facts’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose generation only where output variation is acceptable’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘separate reusable templates from sensitive runtime values’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 29

Before a wider rollout of knowledge, prompt and model strategy, the model strategy group reviews the current evidence. The current solution leaves explicit uncertainty behavior when evidence is insufficient unresolved, and the gap is now affecting the stated business or technical requirement. Which approach best fits the requirement?

1. Set explicit uncertainty behavior when evidence is insufficient.
2. Extend existing Copilot when the workflow stays in Microsoft 365.
3. Define review gates for shared prompt publication.
4. Consider a custom model for a persistent task-specific quality gap.
5. Evaluate resource constraints in an edge scenario.

Correct Answer: A

 

Correct Answer

Answer A is correct because This is correct because it applies the reserved architecture decision—set explicit uncertainty behavior when evidence is insufficient—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘extend existing copilot when the workflow stays in microsoft 365’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘define review gates for shared prompt publication’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer D is incorrect because This is not the best choice because it addresses the adjacent decision ‘consider a custom model for a persistent task-specific quality gap’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘evaluate resource constraints in an edge scenario’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

 

Question 30

An earlier design assumed the wrong thing about context to information relevant to the decision. Which decision should replace that assumption?

1. Compare specialized model latency against quality requirements.
2. Retire redundant templates with conflicting behavior.
3. Reject customization because representative training data is absent makes that approach unsuitable.
4. Limit context to information relevant to the decision.
5. Choose a custom agent because orchestration must be controlled is the decisive constraint.

Correct Answer: D

 

Correct Answer

Answer D is correct because This is correct because it applies the reserved architecture decision—limit context to information relevant to the decision—to the decisive constraint in the scenario. It keeps the action at the appropriate knowledge, prompt and model strategy boundary and produces a result that can be verified before wider deployment.

Incorrect Answers

Answer A is incorrect because This is not the best choice because it addresses the adjacent decision ‘compare specialized model latency against quality requirements’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer B is incorrect because This is not the best choice because it addresses the adjacent decision ‘retire redundant templates with conflicting behavior’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer C is incorrect because This is not the best choice because it addresses the adjacent decision ‘reject customization when representative training data is absent’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.

Answer E is incorrect because This is not the best choice because it addresses the adjacent decision ‘choose a custom agent when orchestration must be controlled’ rather than the scenario’s decisive requirement. Although that action can be valid within knowledge, prompt and model strategy, selecting it here would leave the stated constraint unresolved or move the design to the wrong stage.