Anthropic CCAR-P (Claude Certified Architect - Professional) Practice Test 86

Question 850

Following a regression in human-resources services, the team establishes that a single literal query can miss synonym-heavy material. Which proposed redesign most directly prevents recurrence?

  1. Apply Exact database query through a typed tool
  2. Apply Multi-query expansion for varied wording
  3. Apply Relevance-threshold abstention
  4. Apply Graph traversal for relationships

Explanation

Correct Answer

B:

The reported failure calls for the control under which the integration can create bounded alternate queries and merge deduplicated matches. A prompt-only reminder or unrelated metric would not address the risk that a single literal query can miss synonym-heavy material. Evidence of resolution requires synonym inventory and a query expansion budget and an improvement in recall gain versus extra retrieval calls.

Incorrect Answers

A:

This addresses the failure where generative similarity search is unsuitable for exact transactional state, because the integration can retrieve authorized structured records using parameterized queries; yet it would not correct the observed weakness: a single literal query can miss synonym-heavy material.

C:

This addresses the failure where selecting the nearest vector does not make it relevant, because the integration can require evidence strength before generating a confident factual answer; whereas the human-resources services defect follows the path in which a single literal query can miss synonym-heavy material.

D:

When vector neighbors alone do not prove multi-hop entity relationships, the integration can traverse an authorized graph rather than rely on plain passage similarity; yet it would not correct the observed weakness: a single literal query can miss synonym-heavy material.

Question 851

A compliance assurance architecture group is designing a Claude-supported workflow because a large modernization analysis overwhelms a single unconstrained prompt. Which architectural decision most directly meets that requirement?

  1. Adopt Map-reduce summarization
  2. Adopt Hierarchical work-breakdown structure
  3. Adopt Exception-first task splitting
  4. Adopt Dependency graph scheduling

Explanation

Correct Answer

B:

The mechanism is specific: the architecture can break the outcome into testable components with explicit dependencies. A credible implementation requires named intermediate deliverables and review owners, while its acceptance test monitors completed dependency milestones and recomposition accuracy. The alternative approaches leave the risk that one giant request conceals intermediate failures and unclear ownership.

Incorrect Answers

A:

For a workload where thousands of independent notes exceed practical context budgets, the architecture can summarize bounded chunks then aggregate their structured findings; whereas the compliance assurance case must guard against the outcome that one giant request conceals intermediate failures and unclear ownership.

C:

For a workload where typical requests succeed but rare regulatory exceptions cause serious errors, the architecture can separate normal and exception branches with stronger checks for the latter; the current decision instead depends on a specific capability: the architecture can break the outcome into testable components with explicit dependencies.

D:

For a workload where downstream analyses sometimes run before upstream data cleansing finishes, the architecture can model the task as a DAG and execute nodes only when prerequisites are satisfied; whereas the compliance assurance case must guard against the outcome that one giant request conceals intermediate failures and unclear ownership.

Question 852

The post-incident analysis for logistics coordination concludes that first-in-first-out truncation may discard decisive exceptions. Which remediation targets the relevant failure boundary?

  1. Enforce Memory separation by trust level
  2. Enforce Token-economics telemetry
  3. Enforce Relevance-ranked context packing
  4. Enforce Source-priority truncation policy

Explanation

Correct Answer

D:

The problem lies in the absence of a mechanism where the model configuration can truncate lowest-priority material while protecting essential instructions and citations. Reinstating it, backed by documented context priority policy, limits the exposure that first-in-first-out truncation may discard decisive exceptions; the engineering team should track retained authoritative facts after context trimming.

Incorrect Answers

A:

Where user preferences and external tool outputs enter the same conversation memory, the model configuration can store verified state separately from untrusted retrieved text and time-bound hints; whereas the logistics coordination defect follows the path in which first-in-first-out truncation may discard decisive exceptions.

B:

This addresses the failure where request totals alone cannot isolate token-heavy work, because the model configuration can record uncached input, cache reads, output and tool-related usage per task; yet it would not correct the observed weakness: first-in-first-out truncation may discard decisive exceptions.

C:

Where a case includes more background than the model can productively use, the model configuration can rank excerpts by task relevance and retain traceable source references; but this remedy operates on another boundary, not the case where first-in-first-out truncation may discard decisive exceptions.

Question 853

An operations analyst for research administration must evaluate Pre-generation context pruning on real tasks. A test must detect whether after-the-fact response shortening cannot recover input token spend, even when outputs appear fluent. Which measure is appropriate?

  1. Measure accepted outcome cost by escalation bucket
  2. Measure retained answer support per input token
  3. Measure timeout impact on critical-path latency and grounded-answer rate
  4. Measure finished items per dollar within the batch window

Explanation

Correct Answer

B:

The release should evaluate retained answer support per input token rather than equate successful calls with real outcomes. The control works on the premise that the evaluation process can deduplicate and rank passages before model invocation; the proposed measure can expose the risk that after-the-fact response shortening cannot recover input token spend.

Incorrect Answers

A:

When lightweight answers work on simple cases but fail specialized regulatory reasoning, the evaluation process can raise model capacity selectively after evidence-based difficulty checks; but the acceptance gate must instead inspect retained answer support per input token to detect the targeted failure.

C:

When a task waits for a slow optional analytics integration, the evaluation process can use per-dependency deadlines and continue only if required evidence is available; however, that proxy would not demonstrate retained answer support per input token for the selected intervention.

D:

The alternative indicator checks whether the evaluation process can group asynchronous requests using the available batch workflow when offline processing runs thousands of independent requests nightly; but the acceptance gate must instead inspect retained answer support per input token to detect the targeted failure.

Question 854

The travel-support service has architecture diagrams and alerts, but on-call engineers have no incident, escalation, or recovery procedure. Which operating risk does the missing runbook create?

  1. An exposure in which an architecture overview does not tell operators what to do at 3 a.m.
  2. An exposure in which source-control of application code alone does not recreate inference behavior
  3. An exposure in which documentation volume without traceability is hard to audit or change
  4. An exposure in which an unqualified reference diagram cannot prove actual deployment controls
  5. An exposure in which a high-level component box diagram can hide secondary retention paths

Explanation

Correct Answer

A:

This is the direct residual hazard: an architecture overview does not tell operators what to do at 3 a.m. The omitted intervention matters because the delivery team can document symptoms, safe mitigation, escalation and rollback procedures. A reviewer should demand reviewed failure modes and on-call ownership and measure recovery drill success and time to restore service before treating the risk as controlled.

Incorrect Answers

B:

The competing hazard is that source-control of application code alone does not recreate inference behavior in a workload where multiple environments produce divergent behavior despite using one codebase; however, the named safeguard addresses the distinct hazard that an architecture overview does not tell operators what to do at 3 a.m.

C:

The competing hazard is that documentation volume without traceability is hard to audit or change in a workload where stakeholders cannot tell which architectural control satisfies which need; whereas removing Operational runbook primarily exposes the risk that an architecture overview does not tell operators what to do at 3 a.m.

D:

If legal and platform engineers need to know which environments store data, an unqualified reference diagram cannot prove actual deployment controls; but this is not the immediate loss associated with omitting Operational runbook: an architecture overview does not tell operators what to do at 3 a.m.

E:

If privacy reviewers cannot see where source documents or traces travel, a high-level component box diagram can hide secondary retention paths; however, the named safeguard addresses the distinct hazard that an architecture overview does not tell operators what to do at 3 a.m.

Question 855

A media licensing team is selecting the next improvement for its AI service; a user requests deletion of personal data from model-related application stores. What design choice has the strongest direct justification?

  1. Implement Data residency deployment review
  2. Implement FedRAMP authorization boundary check
  3. Implement Retention and erasure workflow
  4. Implement HIPAA responsibility and contract assessment

Explanation

Correct Answer

C:

The architectural distinction is that the control layer can map and process erasure across prompts, logs, caches and derived indexes as applicable. The decision should be implemented with retention schedules and deletion ownership per data store and evaluated against completed deletion requests and residual-copy checks; otherwise, deleting a front-end record may leave copies in downstream storage.

Incorrect Answers

A:

For a workload where contract terms restrict where customer records may be processed or stored, the control layer can verify regional processing, logs, backups and support access as a whole; that does not provide the capability actually needed: the control layer can map and process erasure across prompts, logs, caches and derived indexes as applicable.

B:

When a US federal workload has explicit authorized-environment constraints, the control layer can validate the complete deployment and service boundary against agency requirements; the alternative cannot replace this key requirement: the control layer can map and process erasure across prompts, logs, caches and derived indexes as applicable.

D:

If a US clinical workflow handles protected health information through vendors, the control layer can verify applicable HIPAA obligations and required business associate arrangements; that does not provide the capability actually needed: the control layer can map and process erasure across prompts, logs, caches and derived indexes as applicable.

Question 856

The research administration program needs a control under which the integration can create bounded alternate queries and merge deduplicated matches. Select TWO components of a credible implementation plan: the design decision and what must exist first. Choose TWO.

  1. Track exact-field match rate and permission denial correctness
  2. Implement Multi-query expansion for varied wording
  3. Establish calibration dataset and a reviewed confidence threshold
  4. Establish synonym inventory and a query expansion budget
  5. Implement Graph traversal for relationships

Explanation

Correct Answers

B:

The architecture component is justified by its behavior: the integration can create bounded alternate queries and merge deduplicated matches. It targets the exposure that a single literal query can miss synonym-heavy material, while the complementary selected actions provide readiness and, for a three-part package, evidence.

D:

The prerequisite is synonym inventory and a query expansion budget. It provides the operating preparation for a system in which the integration can create bounded alternate queries and merge deduplicated matches; without this setup, the planned safeguard has not been demonstrated as deployable.

Incorrect Answers

A:

This test would validate a different benefit: the integration can retrieve authorized structured records using parameterized queries; but that indicator does not verify recall gain versus extra retrieval calls for the selected architecture.

C:

The alternative setup enables a different mechanism: the integration can require evidence strength before generating a confident factual answer; whereas this intervention requires synonym inventory and a query expansion budget as its operating precondition.

E:

This design choice is aimed at cases where investigators ask how two entities connect through approved relationships; it is effective because the integration can traverse an authorized graph rather than rely on plain passage similarity. The rollout described is instead governed by the case where relevant policy is difficult to find because employees use many aliases and requires a different primary intervention; however, the intended intervention requires that the integration can create bounded alternate queries and merge deduplicated matches.

Question 857

An enterprise IT team is advancing Hierarchical work-breakdown structure beyond its proof of concept. Engineers expect that the architecture can break the outcome into testable components with explicit dependencies. What evidence of setup readiness should be required first?

  1. Arrange source precedence rules and escalation ownership
  2. Arrange stable chunk boundaries and merge rules
  3. Arrange an exception taxonomy and differentiated routing policy
  4. Arrange named intermediate deliverables and review owners

Explanation

Correct Answer

D:

The proposed design needs named intermediate deliverables and review owners as its foundation. This matters because the architecture can break the outcome into testable components with explicit dependencies; a plausible demonstration without the prerequisite can still leave a case where one giant request conceals intermediate failures and unclear ownership. The relevant check is completed dependency milestones and recomposition accuracy.

Incorrect Answers

A:

For the separate challenge where two authoritative repositories return different effective dates, the architecture can separate conflict detection and resolution from answer drafting; but the production dependency for Hierarchical work-breakdown structure is named intermediate deliverables and review owners.

B:

For the separate challenge where thousands of independent notes exceed practical context budgets, the architecture can summarize bounded chunks then aggregate their structured findings; for this release, the team must first verify named intermediate deliverables and review owners.

C:

Where typical requests succeed but rare regulatory exceptions cause serious errors, the architecture can separate normal and exception branches with stronger checks for the latter; however, the named intervention needs named intermediate deliverables and review owners before its protection can be enforced.

Question 858

A cost-cutting proposal for media licensing eliminates Source-priority truncation policy. Which undesirable consequence would become most credible after that change?

  1. An exposure in which persistent contamination can make old tool content govern new decisions
  2. An exposure in which blindly filling the context window can bury critical evidence
  3. An exposure in which a character-count heuristic can understate model-specific token use
  4. An exposure in which request totals alone cannot isolate token-heavy work
  5. An exposure in which raw replay increases cost and may reintroduce superseded requests
  6. An exposure in which first-in-first-out truncation may discard decisive exceptions

Explanation

Correct Answer

F:

Removing the safeguard exposes a failure in which first-in-first-out truncation may discard decisive exceptions. The architecture normally counters this because the model configuration can truncate lowest-priority material while protecting essential instructions and citations. A compensating design would require documented context priority policy and evidence from retained authoritative facts after context trimming.

Incorrect Answers

A:

The competing hazard is that persistent contamination can make old tool content govern new decisions in a workload where user preferences and external tool outputs enter the same conversation memory; however, the named safeguard addresses the distinct hazard that first-in-first-out truncation may discard decisive exceptions.

B:

If a case includes more background than the model can productively use, blindly filling the context window can bury critical evidence; whereas removing Source-priority truncation policy primarily exposes the risk that first-in-first-out truncation may discard decisive exceptions.

C:

The competing hazard is that a character-count heuristic can understate model-specific token use in a workload where templates unexpectedly exceed request limits after tool results are appended; but this is not the immediate loss associated with omitting Source-priority truncation policy: first-in-first-out truncation may discard decisive exceptions.

D:

If large context traffic causes budget spikes without obvious request-count increases, request totals alone cannot isolate token-heavy work; however, the named safeguard addresses the distinct hazard that first-in-first-out truncation may discard decisive exceptions.

E:

The competing hazard is that raw replay increases cost and may reintroduce superseded requests in a workload where a long case management dialogue continually replays stale turns; whereas removing Source-priority truncation policy primarily exposes the risk that first-in-first-out truncation may discard decisive exceptions.

Question 859

The post-incident analysis for energy operations concludes that after-the-fact response shortening cannot recover input token spend. Which remediation targets the relevant failure boundary?

  1. Introduce Optimization through better task success
  2. Introduce Result-driven model escalation
  3. Introduce Pre-generation context pruning
  4. Introduce Parallelizable workflow optimization

Explanation

Correct Answer

C:

This intervention addresses the failure at its source because the evaluation process can deduplicate and rank passages before model invocation. To keep the issue from recurring, establish relevance evaluation and source-preserving deduplication and verify retained answer support per input token; otherwise after-the-fact response shortening cannot recover input token spend can persist.

Incorrect Answers

A:

When cheapest inference is not cheapest successful business outcome, the evaluation process can compare token savings against full correction and review cost; yet it would not correct the observed weakness: after-the-fact response shortening cannot recover input token spend.

B:

When swapping every request to a premium tier increases cost without matching benefit, the evaluation process can raise model capacity selectively after evidence-based difficulty checks; but this remedy operates on another boundary, not the case where after-the-fact response shortening cannot recover input token spend.

D:

When blindly increasing concurrency can overload downstream quotas, the evaluation process can parallelize work while joining outputs under a bounded concurrency cap; yet it would not correct the observed weakness: after-the-fact response shortening cannot recover input token spend.

Question 860

A research administration team is selecting the next improvement for its AI service; legal and platform engineers need to know which environments store data. What design choice has the strongest direct justification?

  1. Implement Operational runbook
  2. Implement Decision-to-requirement traceability
  3. Implement End-to-end data-flow diagram
  4. Implement Deployment topology and residency record

Explanation

Correct Answer

D:

The architectural distinction is that the delivery team can document region, identity, storage and network paths for each environment. The decision should be implemented with infrastructure inventory and change-control ownership and evaluated against configuration drift from approved deployment record; otherwise, an unqualified reference diagram cannot prove actual deployment controls.

Incorrect Answers

A:

If incident responders rely on the original author for every production failure, the delivery team can document symptoms, safe mitigation, escalation and rollback procedures; whereas the research administration case must guard against the outcome that an unqualified reference diagram cannot prove actual deployment controls.

B:

When stakeholders cannot tell which architectural control satisfies which need, the delivery team can link decisions to acceptance tests and authoritative requirement IDs; that does not provide the capability actually needed: the delivery team can document region, identity, storage and network paths for each environment.

C:

For a workload where privacy reviewers cannot see where source documents or traces travel, the delivery team can show all data stores, transformations, transfers and trust boundaries; the alternative cannot replace this key requirement: the delivery team can document region, identity, storage and network paths for each environment.

Claude Certified Architect - Professional (CCAR-P) Practice Tests