Microsoft AB-100 (Agentic AI Business Solutions Architect) Practice Test 89

Question 880

Operations has not established the cause of this defect: Users receive permission errors after successful package import. The associated setting is: An imported agent can authenticate in dev but not production. Which inspection should precede remediation?

  1. Review backend authorization logs against actor identities
  2. Trace token acquisition and authorization results by environment
  3. Inspect accessible tool actions and underlying API scopes
  4. Audit token use and secret inventory for agent services

Explanation

Correct Answer

B:

For the situation where an imported agent can authenticate in dev but not production, the defect is that users receive permission errors after successful package import; tracing token acquisition and authorization results by environment helps verify the cause-and-effect path: production consent, identity and connection setup are distinct release checks.

Incorrect Answers

A:

In the process where prompt wording persuades the agent to submit a privileged operation, the diagnostic procedure could explain API trusts the model's text assertion of permission; this particular incident instead turns on whether production consent, identity and connection setup are distinct release checks and why users receive permission errors after successful package import.

C:

The evidence this method produces applies to a workflow where an agent should summarize account data but not alter it; it cannot by itself establish why users receive permission errors after successful package import in the current scenario.

D:

Auditing token use and secret inventory for agent services explores a separate connection, where managed identity provides credential-less workload authentication where supported; the diagnosed requirement is an imported agent can authenticate in dev but not production, and the actual defect is users receive permission errors after successful package import.

Question 881

An investment review asks what measurable result would justify continuing an AI solution where network retries can repeat a single purchase request. The original issue is: Two purchase orders appear from one user intent. Which indicator should be tracked?

  1. Duplicate transactions per retried task
  2. Correctly sequenced workflow executions
  3. Unauthorized ancillary actions per task
  4. Escalation rate after repeated unresolved turns

Explanation

Correct Answer

A:

For the business use where network retries can repeat a single purchase request, the result to quantify is duplicate transactions per retried task; it connects measurement to the mechanism where idempotency keys tie repeated submissions to one business operation, not merely to agent traffic.

Incorrect Answers

B:

The premise of this indicator is defined flow steps make sequencing and error paths explicit; the architecture being evaluated instead serves network retries can repeat a single purchase request, and its success is measured by duplicate transactions per retried task.

C:

For the situation where a bot only needs to check warranty and prepare a draft, unauthorized ancillary actions per task is a sensible indicator; the problem that two purchase orders appear from one user intent requires measurement of duplicate transactions per retried task.

D:

The proposed metric concerns contact Center agents sometimes cannot resolve complex billing intent, whereas the failure in this case is that two purchase orders appear from one user intent; the indicator needed here is duplicate transactions per retried task.

Question 882

A pilot scorecard needs evidence of business improvement rather than only invocation counts. The use case is: Management attributes a sales lift to Copilot without a comparison. Known defect: Seasonality produces the same change in unassisted teams. Which KPI should be selected?

  1. Improvement in domain error rate versus approved baseline
  2. Task completion without switching application context
  3. Realized annualized net benefit versus validated baseline
  4. High-impact actions executed without required approval
  5. Incremental outcome lift relative to control

Explanation

Correct Answer

E:

The observed shortcoming is that seasonality produces the same change in unassisted teams; measuring incremental outcome lift relative to control gives business owners a directly relevant indicator for the process where management attributes a sales lift to Copilot without a comparison.

Incorrect Answers

A:

The premise of this indicator is a measured capability gap, appropriate dataset and MLOps plan justify customization; the architecture being evaluated instead serves management attributes a sales lift to Copilot without a comparison, and its success is measured by incremental outcome lift relative to control.

B:

For the situation where users must remain in a proprietary work-order application, task completion without switching application context is a sensible indicator; the problem that seasonality produces the same change in unassisted teams requires measurement of incremental outcome lift relative to control.

C:

That scorecard concerns the activity in which a pilot reports gross labor savings but omits operating expenses and its result of realized annualized net benefit versus validated baseline; the reported shortcoming here is seasonality produces the same change in unassisted teams.

D:

A change in high-impact actions executed without required approval is informative for teams propose that an agent perform both low-risk lookups and financial postings; it does not establish whether the current process achieves incremental outcome lift relative to control.

Question 883

A business stakeholder describes this operating situation: Support leads need resolution and escalation insights over time. Testing instead shows that team measures chat volume but cannot see failed conversations. What design approach is most appropriate?

  1. Create actionable SLOs and incident alerts for agent flows
  2. Use managed solutions for governed downstream environments
  3. Parameterize environment-specific connection references and variables
  4. Use Copilot Studio analytics for conversation outcomes

Explanation

Correct Answer

D:

The mechanism that agent analytics surface outcome trends and usage signals for supported sessions fits the scenario where support leads need resolution and escalation insights over time; it addresses the defect that team measures chat volume but cannot see failed conversations.

Incorrect Answers

A:

For operations must distinguish healthy traffic from quality incidents, this other approach relies on the fact that service-level objectives and alert thresholds turn telemetry into response signals; the issue here is team measures chat volume but cannot see failed conversations; that mechanism does not establish the required agent analytics surface outcome trends and usage signals for supported sessions.

B:

For production components should not be edited casually, this other approach relies on the fact that managed solution deployment supports controlled downstream customization boundaries; the issue here is team measures chat volume but cannot see failed conversations; that mechanism does not establish the required agent analytics surface outcome trends and usage signals for supported sessions.

C:

Agent writes real customer data into the sandbox; the alternate design can address that situation because environment configuration separates deployable logic from endpoint bindings, whereas this question needs a solution to team measures chat volume but cannot see failed conversations.

Question 884

Business leaders reject an activity-only metric for this process: An AI-driven replenishment proposal influences procurement. They need to detect the issue where recommendation changes spread to all planners without evaluation. Which performance indicator fits?

  1. Sensitive output disclosures in model misuse evaluations
  2. Cross-app exceptions ending in consistent state
  3. Forecast-driven recommendation errors during canary rollout
  4. AI-related incidents during financial-close windows

Explanation

Correct Answer

C:

For the business use where an AI-driven replenishment proposal influences procurement, the result to quantify is forecast-driven recommendation errors during canary rollout; it connects measurement to the mechanism where a pilot cohort and rollback path isolate operational impact, not merely to agent traffic.

Incorrect Answers

A:

That scorecard concerns the activity in which an internal model sometimes returns confidential text verbatim and its result of sensitive output disclosures in model misuse evaluations; the reported shortcoming here is recommendation changes spread to all planners without evaluation.

B:

For the workflow where a blocked invoice triggers a hold on related procurement actions, cross-app exceptions ending in consistent state is an actionable measure; it does not establish whether recommendation changes spread to all planners without evaluation in the situation under review.

D:

A change in aI-related incidents during financial-close windows is informative for invoice posting agents affect critical close processes; it does not establish whether the current process achieves forecast-driven recommendation errors during canary rollout.

Question 885

The acceptance board asks how to verify a corrected design. The setting is: An assistant responds fluently but cites unrelated support articles. Last time, surface quality scores appear healthy despite wrong documents. Which scenario should the team execute?

  1. Disable knowledge access and verify safe no-evidence behavior
  2. Simulate expected launch concurrency and validate headroom
  3. Query identical terms as HR and non-HR users
  4. Use an annotated query with a known authoritative paragraph

Explanation

Correct Answer

D:

The historical failure is that surface quality scores appear healthy despite wrong documents; for the setting where an assistant responds fluently but cites unrelated support articles, using an annotated query with a known authoritative paragraph verifies the dependency where retrieval telemetry distinguishes evidence search from answer generation.

Incorrect Answers

A:

Although the exercise might catch the agent confidently answers when retrieval fails, it would not directly validate the mechanism that retrieval telemetry distinguishes evidence search from answer generation in the target implementation.

B:

The tested behavior is that capacity metrics and throttling errors show saturation risk; the necessary evidence here concerns retrieval telemetry distinguishes evidence search from answer generation and the previously observed surface quality scores appear healthy despite wrong documents.

C:

This check exercises a different system behavior: retrieval applies effective source entitlements to candidate documents; a pass there cannot establish that surface quality scores appear healthy despite wrong documents no longer occurs in the required workflow.

Question 886

An agent-assisted service does not behave as specified: The model may try a tool outside the user's rights. It is used in this situation: A user asks the agent to read an unrelated executive record. Which observable diagnostic procedure is best?

  1. Compare abandonment and handoff reasons by channel
  2. Inspect accessible tool actions and underlying API scopes
  3. Compare requested export destinations against policy allowlists
  4. Inspect backend denials and attempted tool arguments

Explanation

Correct Answer

D:

The model may try a tool outside the user's rights; to attribute the defect rather than guess at a fix, inspecting backend denials and attempted tool arguments examines the dependency where authorization evaluations verify server-side deny behavior.

Incorrect Answers

A:

The proposed inspection is associated with chat and voice agents have different interaction patterns, where channel-specific analytics reveal where interactions fail; it would not expose the failure mode that the model may try a tool outside the user's rights with the required specificity.

B:

Retrieval configuration accidentally grants write capability; examining that other path does not determine whether authorization evaluations verify server-side deny behavior, which is what distinguishes the issue in the stated use case.

C:

The evidence this method produces applies to a workflow where a prompt injection requests data export to an outside address; it cannot by itself establish why the model may try a tool outside the user's rights in the current scenario.

Question 887

A staging test has to detect the specific defect that a shared connection identity grants sandbox agents production write access. In the workflow where test automation must not reach live customer records, which exercise verifies the corrective behavior?

  1. Rotate away from a test secret and verify identity-based access
  2. Attempt an unapproved production modification in test governance
  3. Make identical queries as permitted and denied users
  4. Attempt a cross-environment write using staging credentials

Explanation

Correct Answer

D:

The check is targeted rather than generic: environment-scoped identities isolate access and simplify incident containment; its pass/fail evidence applies to the workflow where test automation must not reach live customer records and to the prior defect that a shared connection identity grants sandbox agents production write access.

Incorrect Answers

A:

The tested behavior is that managed identity provides credential-less workload authentication where supported; the necessary evidence here concerns environment-scoped identities isolate access and simplify incident containment and the previously observed a shared connection identity grants sandbox agents production write access.

B:

Although the exercise might catch a maker changes production agent logic outside the release process, it would not directly validate the mechanism that environment-scoped identities isolate access and simplify incident containment in the target implementation.

C:

The verification concerning shared result caching makes answers fast across roles checks cache keys and access validation prevent security-context leakage against one user's cached confidential answer appears for another user; its result says little about the observed issue where a shared connection identity grants sandbox agents production write access.

Question 888

A pilot scorecard needs evidence of business improvement rather than only invocation counts. The use case is: Teams rebuild similar assistants without seeing existing assets. Known defect: Multiple agents offer overlapping tools with inconsistent permissions. Which KPI should be selected?

  1. Percentage of agent releases passing common governance gates
  2. Incremental outcome lift relative to control
  3. Conflicting cross-application writes per thousand transactions
  4. Reuse rate of registered AI components across business units

Explanation

Correct Answer

D:

For the business use where teams rebuild similar assistants without seeing existing assets, the result to quantify is reuse rate of registered AI components across business units; it connects measurement to the mechanism where discovery and ownership metadata support reuse and controlled deprecation, not merely to agent traffic.

Incorrect Answers

A:

For the situation where independent teams choose inconsistent evaluation and security methods, percentage of agent releases passing common governance gates is a sensible indicator; the problem that multiple agents offer overlapping tools with inconsistent permissions requires measurement of reuse rate of registered AI components across business units.

B:

The proposed metric concerns management attributes a sales lift to Copilot without a comparison, whereas the failure in this case is that multiple agents offer overlapping tools with inconsistent permissions; the indicator needed here is reuse rate of registered AI components across business units.

C:

Although conflicting cross-application writes per thousand transactions is observable, the affected process involves teams rebuild similar assistants without seeing existing assets; its acceptance outcome is reuse rate of registered AI components across business units, not the alternative measurement.

Question 889

A pilot scorecard needs evidence of business improvement rather than only invocation counts. The use case is: Users ask about governed CRM records rather than prose manuals. Known defect: A document index cannot answer current case ownership reliably. Which KPI should be selected?

  1. Share of material claims supported by retrievable citations
  2. New-policy answer accuracy after ingestion
  3. Share of failed workflows with an identified responsible component
  4. Authorized structured-record answer accuracy

Explanation

Correct Answer

D:

Authorized structured-record answer accuracy; this outcome measure belongs to the workflow where users ask about governed CRM records rather than prose manuals; a change in it can help detect whether the defect that a document index cannot answer current case ownership reliably is resolved.

Incorrect Answers

A:

The alternative tracks share of material claims supported by retrievable citations, arising from the mechanism where cited retrieval separates factual support from free-form generation; its movement alone says little about the failure that a document index cannot answer current case ownership reliably.

B:

That scorecard concerns the activity in which answers depend on policies revised every week and its result of new-policy answer accuracy after ingestion; the reported shortcoming here is a document index cannot answer current case ownership reliably.

C:

Share of failed workflows with an identified responsible component; that indicator describes a distinct workflow where copilot, Dataverse, Dynamics and custom APIs share one process; it does not quantify authorized structured-record answer accuracy, which is needed for the stated case.

Question 890

A solution architecture is being reviewed for this business situation: A regulated workload must keep data in permitted locations. The observed defect is that an external model call exports sensitive prompts to an unapproved region. Which architectural decision addresses the underlying requirement?

  1. Treat index schema and embeddings as versioned artifacts
  2. Map all agent data flows to approved geographic boundaries
  3. Apply data loss prevention policies to connector combinations
  4. Record immutable model deployment events with approver and version

Explanation

Correct Answer

B:

A regulated workload must keep data in permitted locations; its failing behavior is that an external model call exports sensitive prompts to an unapproved region; applying the underlying design principle—data-flow architecture identifies processing, storage and telemetry locations—connects the requirement to an implementable control.

Incorrect Answers

A:

The alternative addresses existing vectors are queried with incompatible assumptions by a design in which explicit index versions support compatible retrieval and rollback; it does not supply the missing relationship for a regulated workload must keep data in permitted locations, where an external model call exports sensitive prompts to an unapproved region.

C:

In the case where a Copilot Studio agent can read corporate data and use public messaging, the alternate design depends on Power Platform data policies constrain connector use and movement pathways; it does not correct the target failure, an external model call exports sensitive prompts to an unapproved region.

D:

In the case where a model is replaced without a traceable change ticket, the alternate design depends on deployment audit links actor, approved artifact and release time; it does not correct the target failure, an external model call exports sensitive prompts to an unapproved region.

Agentic AI Business Solutions Architect (AB-100) Practice Tests