Question 380
After deployment, stakeholders need an interpretable measure for this use case: Independent teams choose inconsistent evaluation and security methods. The pilot showed that one division releases agents without risk review. Which metric captures the needed result?
- Interoperable tasks completed without custom point-to-point parsing
- Fully loaded cost per correctly resolved case
- Percentage of agent releases passing common governance gates
- Reuse rate of registered AI components across business units
Explanation
Correct Answer
C:The observed shortcoming is that one division releases agents without risk review; measuring percentage of agent releases passing common governance gates gives business owners a directly relevant indicator for the process where independent teams choose inconsistent evaluation and security methods.
Incorrect Answers
A:A change in interoperable tasks completed without custom point-to-point parsing is informative for independently hosted agents must exchange tasks and results; it does not establish whether the current process achieves percentage of agent releases passing common governance gates.
B:That scorecard concerns the activity in which AI drafts requests quickly but specialists must correct some outputs and its result of fully loaded cost per correctly resolved case; the reported shortcoming here is one division releases agents without risk review.
D:The alternative tracks reuse rate of registered AI components across business units, arising from the mechanism where discovery and ownership metadata support reuse and controlled deprecation; its movement alone says little about the failure that one division releases agents without risk review.
Question 381
An enterprise team is making a platform design decision. Business context: A research agent repeatedly queries sources on difficult cases. Present failure: A single case exhausts model and tool quotas. Which approach resolves the mismatch?
- Specify stop conditions and budgets for long-running agents
- Trigger an autonomous agent from a governed business event
- Disambiguate overlapping topic descriptions and triggers
- Create a curated evaluation set for retrieval quality
Explanation
Correct Answer
A:For the need that a research agent repeatedly queries sources on difficult cases, the relevant mechanism is that time, step and spend budgets terminate unproductive loops; that architectural dependency addresses the observed failure where a single case exhausts model and tool quotas.
Incorrect Answers
B:When shipment exceptions should initiate investigation without a chat message, an event trigger starts the bounded agent workflow on state change; in the stated workflow, however, a single case exhausts model and tool quotas; the relevant architectural behavior is that time, step and spend budgets terminate unproductive loops.
C:Agent invokes the wrong booking action for a reschedule; the alternate design can address that situation because distinct descriptions and inputs improve topic selection, whereas this question needs a solution to a single case exhausts model and tool quotas.
D:The technical behavior of this alternative is that labeled queries and expected evidence distinguish retrieval from synthesis; the user case being assessed is a research agent repeatedly queries sources on difficult cases with a different failure: a single case exhausts model and tool quotas.
Question 382
A quality engineer is writing a regression case after the incident where timeout creates an abandoned workflow with no owner. The underlying business scenario is: An autonomous agent relies on slow approvals and delayed tool results. What should the test exercise?
- Run red-team and representative workflow tests before publishing
- Delay an approver past the defined execution deadline
- Alternate ambiguous and corrected identifiers within one session
- Test a conversation that ends politely without resolving the case
Explanation
Correct Answer
B:The check is targeted rather than generic: state-machine tests examine retries, expiry and compensation; its pass/fail evidence applies to the workflow where an autonomous agent relies on slow approvals and delayed tool results and to the prior defect that timeout creates an abandoned workflow with no owner.
Incorrect Answers
A:The verification concerning a hosted agent gains an additional research tool checks evaluation suites validate quality and safety before publication against new planning behavior occasionally leaks internal notes; its result says little about the observed issue where timeout creates an abandoned workflow with no owner.
C:In the alternate setting where agent must remember order context while asking for missing details, the test exercises multi-turn tests assess state tracking, entity separation and follow-up questions; even a pass would not show that timeout creates an abandoned workflow with no owner has been prevented.
D:The tested behavior is that outcome measures connect agent use with completed business work; the necessary evidence here concerns state-machine tests examine retries, expiry and compensation and the previously observed timeout creates an abandoned workflow with no owner.
Question 383
The fault affects an AI-powered workflow in which employees install external agents with unknown data pathways. Specifically, an unreviewed plugin exports sensitive tenant information. Which diagnostic investigation best distinguishes the relevant failure path?
- Audit training-data writers and model approvers
- Inspect third-party manifests and requested access scopes
- Review data processing agreements and network endpoints
- Audit dataset readers and movement into training jobs
Explanation
Correct Answer
B:Where an unreviewed plugin exports sensitive tenant information, the useful evidence comes from inspecting third-party manifests and requested access scopes; its focus matches the architectural mechanism that procurement and security review evaluate permissions and data flows in the process where employees install external agents with unknown data pathways.
Incorrect Answers
A:Unvetted labeled data poisons a retrained classifier; examining that other path does not determine whether procurement and security review evaluate permissions and data flows, which is what distinguishes the issue in the stated use case.
C:That probe evaluates contract and privacy assessment cover service location and subprocessor use in a workflow where a third-party AI tool changes where submitted documents are processed; the reported failure concerns the path for which procurement and security review evaluate permissions and data flows instead.
D:Training examples become accessible outside intended teams; examining that other path does not determine whether procurement and security review evaluate permissions and data flows, which is what distinguishes the issue in the stated use case.
Question 384
The rollout supports this workflow: A service bot handles many chats but few cases are resolved. Leaders want a metric that responds to the reported shortcoming that a dashboard calls the bot successful because messages increased. Which measure best reflects success?
- Total processing cost per correct verified extraction
- Orphaned business workflows after simulated timeouts
- Duplicate side effects per one thousand actions
- Verified task resolution rate
Explanation
Correct Answer
D:The use case involves a service bot handles many chats but few cases are resolved and the baseline failure that a dashboard calls the bot successful because messages increased; the KPI tracks verified task resolution rate, so improvement has an interpretable operational meaning.
Incorrect Answers
A:Although total processing cost per correct verified extraction is observable, the affected process involves a service bot handles many chats but few cases are resolved; its acceptance outcome is verified task resolution rate, not the alternative measurement.
B:For the situation where an autonomous agent relies on slow approvals and delayed tool results, orphaned business workflows after simulated timeouts is a sensible indicator; the problem that a dashboard calls the bot successful because messages increased requires measurement of verified task resolution rate.
C:The proposed metric concerns agents modify records through external APIs, whereas the failure in this case is that a dashboard calls the bot successful because messages increased; the indicator needed here is verified task resolution rate.
Question 385
Before production promotion, the team needs evidence for this setting: Agent can research options but cannot approve a vendor payment. A prior regression caused the following: A background process posts expenses beyond its delegated scope. Which validation is most persuasive?
- Run representative notes and validate schema and exclusions
- Drive a recommendation above the threshold and confirm blocking
- Test edge cases with conflicting documents and incomplete facts
- Change a test case status and inspect audit events
Explanation
Correct Answer
B:The check is targeted rather than generic: approval checkpoints split independent analysis from high-impact execution; its pass/fail evidence applies to the workflow where agent can research options but cannot approve a vendor payment and to the prior defect that a background process posts expenses beyond its delegated scope.
Incorrect Answers
A:This check exercises a different system behavior: prompt actions encapsulate reusable inputs and expected outputs; a pass there cannot establish that a background process posts expenses beyond its delegated scope no longer occurs in the required workflow.
C:In the alternate setting where procurement exceptions require comparing several policies and records, the test exercises reasoning can plan through evidence and tools within enforced boundaries; even a pass would not show that a background process posts expenses beyond its delegated scope has been prevented.
D:Changing a test case status and inspecting audit events verifies an adjacent objective where an agent needs to update a governed ticket status; it leaves untested the causal path in which approval checkpoints split independent analysis from high-impact execution for this case.
Question 386
Troubleshooting starts from a business complaint: A model retrieves a draft manual without any accountable owner. The affected process is: Search spans internal manuals with different owners and permissions. Which investigation is most directly tied to the reported defect?
- Audit samples for missing owner, classification and source identifiers
- Compare retrieved versions against the policy registry
- Compare retrieval results for users with different permissions
- Trace transactions using end-to-end correlation IDs
Explanation
Correct Answer
A:In the setting where search spans internal manuals with different owners and permissions, auditing samples for missing owner, classification and source identifiers can isolate the failure in which a model retrieves a draft manual without any accountable owner, because metadata enables filtering, review, retention and traceable answers.
Incorrect Answers
B:That probe evaluates version and effective-date metadata select authoritative content at retrieval time in a workflow where policy documents contain draft, superseded, and currently effective revisions; the reported failure concerns the path for which metadata enables filtering, review, retention and traceable answers instead.
C:The proposed inspection is associated with employee policies reside in governed Microsoft 365 documents, where supported SharePoint grounding retrieves organizational content within access controls; it would not expose the failure mode that a model retrieves a draft manual without any accountable owner with the required specificity.
D:Copilot, Dataverse, Dynamics and custom APIs share one process; its matching diagnostic check is reasonable for failures cannot be attributed to any single integration owner; the present investigation needs evidence for the different failure where a model retrieves a draft manual without any accountable owner.
Question 387
A design adds autonomous behavior to a process where a request crosses Copilot Studio, a flow and the CRM backend. This creates a documented risk: Correlation identifiers might embed patient or customer details as stable identifiers. Which enforced control is the best fit?
- Do not approve procurement while financial hold status is unknown
- Avoid recording personal identifiers as correlation keys
- Use idempotency on any repeatable mutation action
- Block release when production variables or connections are unresolved
- Roll back or withhold publication if acceptance checks fail
Explanation
Correct Answer
B:Because correlation identifiers might embed patient or customer details as stable identifiers, a guardrail must constrain the actual system behavior: correlation identifiers link the distributed operational path; protecting a response alone cannot enforce that boundary.
Incorrect Answers
A:A blocked invoice triggers a hold on related procurement actions; its guardrail addresses the possibility that a procurement approval might proceed before a finance hold is resolved; the present architecture must constrain the different threat that correlation identifiers might embed patient or customer details as stable identifiers.
C:This restriction is associated with agents modify records through external APIs and addresses the exposure that retries might cause repeated transactions even though tool calls appear successful; it does not specifically protect the target workflow against correlation identifiers might embed patient or customer details as stable identifiers.
D:The alternate guard addresses imported actions might retain staging endpoints after production deployment for a flow moved to production still points to a test endpoint; the present threat, correlation identifiers might embed patient or customer details as stable identifiers, needs a different enforcement boundary.
E:The alternate boundary mitigates an imported but unpublished agent might cause users to see stale behavior, which is relevant to solution import succeeds but users still see the previous behavior; it leaves the current workflow vulnerable to correlation identifiers might embed patient or customer details as stable identifiers.
Question 388
The acceptance board asks how to verify a corrected design. The setting is: A workflow needs proprietary scheduling, model choice and runtime control. Last time, a predefined host cannot meet the required orchestration contract. Which scenario should the team execute?
- Run a representative extraction holdout at production concurrency
- Prototype a unique scheduling dependency with the custom runtime
- Call a read-only tool with valid and invalid identity contexts
- Replace a source document without training and verify updated answers
Explanation
Correct Answer
B:A predefined host cannot meet the required orchestration contract; a release test must establish that this no longer occurs; prototyping a unique scheduling dependency with the custom runtime measures the specific behavior for which owning the engine provides control over models, routing and execution.
Incorrect Answers
A:For high-volume narrow extraction needs predictable latency and cost, the alternate test is about whether smaller models can be evaluated for narrow tasks with lower cost; it cannot prove the outcome needed after the reported regression where a predefined host cannot meet the required orchestration contract.
C:A pass on that verification would speak to a Copilot Studio agent needs a reusable interface to governed operations; it would not show that owning the engine provides control over models, routing and execution, which is needed to avoid the observed defect a predefined host cannot meet the required orchestration contract.
D:Although the exercise might catch a trained model remembers obsolete guidance after source changes, it would not directly validate the mechanism that owning the engine provides control over models, routing and execution in the target implementation.
Question 389
Troubleshooting starts from a business complaint: Universal use of a premium model inflates cost. The affected process is: Many calls require only a small fixed-label output. Which investigation is most directly tied to the reported defect?
- Compare host limitations against the requirements matrix
- Score use cases against data, identity, governance and integration readiness
- Compare orchestration requirements with host extensibility limits
- Analyze route decisions by task class and observed error
Explanation
Correct Answer
D:Universal use of a premium model inflates cost; to attribute the defect rather than guess at a fix, analyzing route decisions by task class and observed error examines the dependency where task-aware routing chooses economical models after quality gates.
Incorrect Answers
A:A predefined host cannot meet the required orchestration contract; examining that other path does not determine whether task-aware routing chooses economical models after quality gates, which is what distinguishes the issue in the stated use case.
B:In the separate scenario where candidate processes are attractive but permissions and data quality are weak, the same investigation could explain the first-wave pilot requests broad access to unreliable records; it does not isolate the reported cause in a system where many calls require only a small fixed-label output.
C:The evidence this method produces applies to a workflow where the solution demands its own model selection and hosting policies; it cannot by itself establish why universal use of a premium model inflates cost in the current scenario.
Question 390
The pilot needs a durable correction: Retrieval configuration accidentally grants write capability. This affects the workflow in which an agent should summarize account data but not alter it. Which design choice should the architect approve?
- Separate read-only grounding from transactional tools
- Test tool-permission escalation attempts at the execution layer
- Require independent backend authorization for generated tool calls
- Test posting permissions and approvals after agent updates
Explanation
Correct Answer
A:For the need that an agent should summarize account data but not alter it, the relevant mechanism is that distinct permission surfaces enforce least privilege by task; that architectural dependency addresses the observed failure where retrieval configuration accidentally grants write capability.
Incorrect Answers
B:Authorization evaluations verify server-side deny behavior; its main purpose is a user asks the agent to read an unrelated executive record, not correction of the reported defect where retrieval configuration accidentally grants write capability in the target process.
C:For another type of architecture, server-side access checks enforce real actor entitlements is useful; this design must instead support an agent should summarize account data but not alter it despite the problem where retrieval configuration accidentally grants write capability.
D:There is a valid place for the mechanism that authorization regression gates verify posting authority remains constrained, namely agents suggest payment adjustments and inventory holds; it does not remove the specific defect that retrieval configuration accidentally grants write capability from this case.