Anthropic CCAR-P (Claude Certified Architect - Professional) Practice Test 8

Question 70

A parallel specialist pipeline appears fast, but its supposedly independent tasks modify the same shared object. Which result proves the requested control, not just configuration?

  1. A production-like test shows that injected sum errors are rejected by the calculation stage before invoices enter approval.
  2. The release check confirms that critical-path tracing identifies slow steps and confirms nondependent checks run concurrently.
  3. The reviewer observes that each stage has its own labeled evaluation, with missing clauses isolated before drafting.
  4. The release report verifies that every material incident conclusion maps to a retrieved log or reviewed artifact.
  5. The controlled replay verifies that a disputed refund follows encoded policy while tone variation changes only the explanation.

Explanation

Correct Answer

B:

Explicit dependency analysis avoids serialization when subtasks can be safely evaluated independently; the test is necessary where a complex workflow spans independent specialist checks with different latency targets.

Incorrect Answers

A:

Arithmetic should be checked by exact computation rather than relying on probabilistic text generation; this makes sense when an invoice assistant must verify totals exactly after interpreting handwritten line items, rather than the risk that a parallel specialist pipeline appears fast, but its supposedly independent tasks modify the same shared object.

C:

Distinct intermediate outputs make omissions diagnosable and reduce conflicting instructions in one model call; the trigger would be that a contract review prompt misses exceptions because it tries to extract, assess and draft at once, while the immediate question involves a parallel specialist pipeline appears fast, but its supposedly independent tasks modify the same shared object.

D:

Evidence-first decomposition makes claim attribution explicit instead of asking a final model to backfill sources; the design helps when an incident report requires citations for every claim but one final prompt invents them, but the present constraint is that a parallel specialist pipeline appears fast, but its supposedly independent tasks modify the same shared object.

E:

Separating authority from language generation keeps business rules outside model discretion; that fits cases where a user request combines a refund decision and a sentiment summary, but only the latter is model-suitable, whereas the current issue is that a parallel specialist pipeline appears fast, but its supposedly independent tasks modify the same shared object.

Question 71

A sales assistant beats baseline accuracy but costs three times the approved cost per transaction. How should the architect revise the proposed solution?

  1. In the implementation sequence, use asynchronous batching and throughput-oriented scheduling where platform support allows.
  2. For the operating team, compare quality-adjusted unit economics against the agreed per-case cost ceiling.
  3. For this deployment, design and load-test against tail latency rather than average response time.
  4. As a scoped first step, establish baseline labor time and evaluate avoided work plus human-review effort.
  5. In the release design, define a graceful-degradation route and measurable recovery objectives for essential tasks.

Explanation

Correct Answer

B:

Accuracy improvement alone does not establish feasibility when the deployed workload violates unit economics; a meaningful test result is that a measured cost-per-accepted-response report includes inference, retrieval and review expenses.

Incorrect Answers

A:

Batching can reduce per-item cost when results are not needed during an active user session; the other workload is one where a bulk overnight document run does not require interactive latency, not the risk where a sales assistant beats baseline accuracy but costs three times the approved cost per transaction.

C:

Rare slow requests determine user-facing SLA risk that median timing can conceal; the alternative helps if a contact center needs 95th-percentile responses under its contractual target, whereas this scenario involves a sales assistant beats baseline accuracy but costs three times the approved cost per transaction.

D:

Credible ROI includes both productivity gains and the operational work needed to make outputs trustworthy; it is justified if a finance sponsor asks for an ROI claim before any control group has been measured, but here a sales assistant beats baseline accuracy but costs three times the approved cost per transaction.

E:

An application SLA depends on fallback behavior and failure isolation, not just model service uptime; its scope includes cases where a rollout must keep critical tasks available when the model endpoint is impaired, instead of when a sales assistant beats baseline accuracy but costs three times the approved cost per transaction.

Question 72

A contact center needs 95th-percentile responses under its contractual target. Which control should govern the first implementation?

  1. Under the system plan, define a graceful-degradation route and measurable recovery objectives for essential tasks.
  2. As the design priority, design and load-test against tail latency rather than average response time.
  3. In the implementation sequence, set tool-call budgets and benchmark whether simpler bounded workflows meet quality targets.
  4. For the operating team, establish baseline labor time and evaluate avoided work plus human-review effort.

Explanation

Correct Answer

B:

Rare slow requests determine user-facing SLA risk that median timing can conceal; a representative trial should prove that a production-like load test meets p95 response time at expected peak concurrency.

Incorrect Answers

A:

An application SLA depends on fallback behavior and failure isolation, not just model service uptime; it serves conditions where a rollout must keep critical tasks available when the model endpoint is impaired, but this decision concerns a contact center needs 95th-percentile responses under its contractual target.

C:

Additional steps are worthwhile only if their incremental quality benefit offsets latency and cost; this makes sense when a proposed agent gains marginal accuracy while invoking tools dozens of times per ticket, but the requested boundary concerns a contact center needs 95th-percentile responses under its contractual target.

D:

Credible ROI includes both productivity gains and the operational work needed to make outputs trustworthy; this technique fits where a finance sponsor asks for an ROI claim before any control group has been measured, not the present case where a contact center needs 95th-percentile responses under its contractual target.

Question 73

A bulk overnight document run does not require interactive latency. What intervention best matches the stated boundary?

  1. For the next architecture revision, design and load-test against tail latency rather than average response time.
  2. For the current integration, define a graceful-degradation route and measurable recovery objectives for essential tasks.
  3. Under the system plan, establish baseline labor time and evaluate avoided work plus human-review effort.
  4. As the design priority, use asynchronous batching and throughput-oriented scheduling where platform support allows.

Explanation

Correct Answer

D:

Batching can reduce per-item cost when results are not needed during an active user session; the go-live check should establish that completed documents arrive within the overnight window at lower measured unit cost.

Incorrect Answers

A:

Rare slow requests determine user-facing SLA risk that median timing can conceal; it serves a setting where a contact center needs 95th-percentile responses under its contractual target, rather than the condition that a bulk overnight document run does not require interactive latency.

B:

An application SLA depends on fallback behavior and failure isolation, not just model service uptime; that is defensible if a rollout must keep critical tasks available when the model endpoint is impaired, while the immediate question involves a bulk overnight document run does not require interactive latency.

C:

Credible ROI includes both productivity gains and the operational work needed to make outputs trustworthy; the other workload is one where a finance sponsor asks for an ROI claim before any control group has been measured, but the present constraint is that a bulk overnight document run does not require interactive latency.

Question 74

A proposed agent gains marginal accuracy while invoking tools dozens of times per ticket. Which implementation decision is the strongest fit?

  1. In the project scope, use asynchronous batching and throughput-oriented scheduling where platform support allows.
  2. Within the rollout design, design and load-test against tail latency rather than average response time.
  3. For the next architecture revision, set tool-call budgets and benchmark whether simpler bounded workflows meet quality targets.
  4. For the current integration, establish baseline labor time and evaluate avoided work plus human-review effort.

Explanation

Correct Answer

C:

Additional steps are worthwhile only if their incremental quality benefit offsets latency and cost; a suitable release result is that the selected design hits acceptance accuracy with a documented maximum number of tool calls.

Incorrect Answers

A:

Batching can reduce per-item cost when results are not needed during an active user session; that belongs to situations where a bulk overnight document run does not require interactive latency, not this situation where a proposed agent gains marginal accuracy while invoking tools dozens of times per ticket.

B:

Rare slow requests determine user-facing SLA risk that median timing can conceal; it is relevant if a contact center needs 95th-percentile responses under its contractual target, whereas this scenario involves a proposed agent gains marginal accuracy while invoking tools dozens of times per ticket.

D:

Credible ROI includes both productivity gains and the operational work needed to make outputs trustworthy; the other case is when a finance sponsor asks for an ROI claim before any control group has been measured, instead of when a proposed agent gains marginal accuracy while invoking tools dozens of times per ticket.

Question 75

A finance sponsor asks for an ROI claim before any control group has been measured. What should be prioritized during architecture review?

  1. For this workflow, include region-specific latency, data-location constraints and availability targets in selection criteria.
  2. For the affected service, use asynchronous batching and throughput-oriented scheduling where platform support allows.
  3. In the project scope, define a graceful-degradation route and measurable recovery objectives for essential tasks.
  4. Within the rollout design, establish baseline labor time and evaluate avoided work plus human-review effort.

Explanation

Correct Answer

D:

Credible ROI includes both productivity gains and the operational work needed to make outputs trustworthy; the practical acceptance condition is that time studies quantify net staff minutes saved per approved case versus the current process.

Incorrect Answers

A:

Deployment economics and compliance are multi-dimensional; cheapest inference alone cannot choose a hosting route; this control is needed when an application serves users across regions with varying response times and residency duties, but the actual risk is that a finance sponsor asks for an ROI claim before any control group has been measured.

B:

Batching can reduce per-item cost when results are not needed during an active user session; that would be useful if a bulk overnight document run does not require interactive latency, but the requested boundary concerns a finance sponsor asks for an ROI claim before any control group has been measured.

C:

An application SLA depends on fallback behavior and failure isolation, not just model service uptime; it serves a setting where a rollout must keep critical tasks available when the model endpoint is impaired, not the present case where a finance sponsor asks for an ROI claim before any control group has been measured.

Question 76

An application serves users across regions with varying response times and residency duties. Which response follows from the actual operational constraint?

  1. At the system boundary, set tool-call budgets and benchmark whether simpler bounded workflows meet quality targets.
  2. For the primary workstream, define a graceful-degradation route and measurable recovery objectives for essential tasks.
  3. In the proposed design, include region-specific latency, data-location constraints and availability targets in selection criteria.
  4. For the affected service, design and load-test against tail latency rather than average response time.
  5. For this production path, use asynchronous batching and throughput-oriented scheduling where platform support allows.

Explanation

Correct Answer

C:

Deployment economics and compliance are multi-dimensional; cheapest inference alone cannot choose a hosting route; the go-live check should establish that tests show each region meets its approved data boundary and responsiveness commitments.

Incorrect Answers

A:

Additional steps are worthwhile only if their incremental quality benefit offsets latency and cost; it targets cases where a proposed agent gains marginal accuracy while invoking tools dozens of times per ticket, while the immediate question involves an application serves users across regions with varying response times and residency duties.

B:

An application SLA depends on fallback behavior and failure isolation, not just model service uptime; the approach helps when a rollout must keep critical tasks available when the model endpoint is impaired, but the present constraint is that an application serves users across regions with varying response times and residency duties.

D:

Rare slow requests determine user-facing SLA risk that median timing can conceal; it is relevant if a contact center needs 95th-percentile responses under its contractual target, whereas the release criterion addresses an application serves users across regions with varying response times and residency duties.

E:

Batching can reduce per-item cost when results are not needed during an active user session; it serves conditions where a bulk overnight document run does not require interactive latency, while the required safeguard concerns an application serves users across regions with varying response times and residency duties.

Question 77

A rollout must keep critical tasks available when the model endpoint is impaired. Which change targets the constraint rather than a proxy metric?

  1. For the intended users, compare quality-adjusted unit economics against the agreed per-case cost ceiling.
  2. Within the delivery plan, include region-specific latency, data-location constraints and availability targets in selection criteria.
  3. At the system boundary, establish baseline labor time and evaluate avoided work plus human-review effort.
  4. In the initial solution, define a graceful-degradation route and measurable recovery objectives for essential tasks.

Explanation

Correct Answer

D:

An application SLA depends on fallback behavior and failure isolation, not just model service uptime; one operational check should confirm that a fault-injection test preserves essential manual processing within recovery targets.

Incorrect Answers

A:

Accuracy improvement alone does not establish feasibility when the deployed workload violates unit economics; the design helps when a sales assistant beats baseline accuracy but costs three times the approved cost per transaction, whereas this scenario involves a rollout must keep critical tasks available when the model endpoint is impaired.

B:

Deployment economics and compliance are multi-dimensional; cheapest inference alone cannot choose a hosting route; that fits cases where an application serves users across regions with varying response times and residency duties, but here a rollout must keep critical tasks available when the model endpoint is impaired.

C:

Credible ROI includes both productivity gains and the operational work needed to make outputs trustworthy; this control is needed when a finance sponsor asks for an ROI claim before any control group has been measured, instead of when a rollout must keep critical tasks available when the model endpoint is impaired.

Question 78

A purchase committee needs the actual cost of approved AI outputs, including failed calls and reviewer minutes. Which measurement is the most relevant release criterion?

  1. The pilot evaluation finds that the selected design hits acceptance accuracy with a documented maximum number of tool calls.
  2. The comparison run confirms that a measured cost-per-accepted-response report includes inference, retrieval and review expenses.
  3. The staged rollout demonstrates that a fault-injection test preserves essential manual processing within recovery targets.
  4. The audit sample demonstrates that completed documents arrive within the overnight window at lower measured unit cost.
  5. The readiness assessment shows that time studies quantify net staff minutes saved per approved case versus the current process.

Explanation

Correct Answer

B:

Accuracy improvement alone does not establish feasibility when the deployed workload violates unit economics; this is relevant for cases where a sales assistant beats baseline accuracy but costs three times the approved cost per transaction.

Incorrect Answers

A:

Additional steps are worthwhile only if their incremental quality benefit offsets latency and cost; it is justified if a proposed agent gains marginal accuracy while invoking tools dozens of times per ticket, but the requested boundary concerns a purchase committee needs the actual cost of approved AI outputs, including failed calls and reviewer minutes.

C:

An application SLA depends on fallback behavior and failure isolation, not just model service uptime; it targets cases where a rollout must keep critical tasks available when the model endpoint is impaired, while this problem involves a purchase committee needs the actual cost of approved AI outputs, including failed calls and reviewer minutes.

D:

Batching can reduce per-item cost when results are not needed during an active user session; the approach helps when a bulk overnight document run does not require interactive latency, and not where a purchase committee needs the actual cost of approved AI outputs, including failed calls and reviewer minutes.

E:

Credible ROI includes both productivity gains and the operational work needed to make outputs trustworthy; that belongs to situations where a finance sponsor asks for an ROI claim before any control group has been measured, rather than the risk that a purchase committee needs the actual cost of approved AI outputs, including failed calls and reviewer minutes.

Question 79

Call-center users report sporadic ten-second waits despite acceptable mean latency. Which production-like test finding is most convincing?

  1. The sign-off record demonstrates that a production-like load test meets p95 response time at expected peak concurrency.
  2. The acceptance evidence shows that tests show each region meets its approved data boundary and responsiveness commitments.
  3. The pilot evaluation finds that time studies quantify net staff minutes saved per approved case versus the current process.
  4. The comparison run confirms that the selected design hits acceptance accuracy with a documented maximum number of tool calls.

Explanation

Correct Answer

A:

Rare slow requests determine user-facing SLA risk that median timing can conceal; the safeguard must work when a contact center needs 95th-percentile responses under its contractual target.

Incorrect Answers

B:

Deployment economics and compliance are multi-dimensional; cheapest inference alone cannot choose a hosting route; the trigger would be that an application serves users across regions with varying response times and residency duties, whereas the current issue is that call-center users report sporadic ten-second waits despite acceptable mean latency.

C:

Credible ROI includes both productivity gains and the operational work needed to make outputs trustworthy; the design helps when a finance sponsor asks for an ROI claim before any control group has been measured, whereas the release criterion addresses call-center users report sporadic ten-second waits despite acceptable mean latency.

D:

Additional steps are worthwhile only if their incremental quality benefit offsets latency and cost; that fits cases where a proposed agent gains marginal accuracy while invoking tools dozens of times per ticket, while the required safeguard concerns call-center users report sporadic ten-second waits despite acceptable mean latency.

Question 80

A noninteractive document service runs overnight but consumes the premium synchronous request budget. Which finding should the release gate require?

  1. A reproducible run shows that a fault-injection test preserves essential manual processing within recovery targets.
  2. The signed test record confirms that completed documents arrive within the overnight window at lower measured unit cost.
  3. A production-like test shows that a production-like load test meets p95 response time at expected peak concurrency.
  4. The acceptance evidence shows that a measured cost-per-accepted-response report includes inference, retrieval and review expenses.

Explanation

Correct Answer

B:

Batching can reduce per-item cost when results are not needed during an active user session; its finding speaks to the case where a bulk overnight document run does not require interactive latency.

Incorrect Answers

A:

An application SLA depends on fallback behavior and failure isolation, not just model service uptime; this applies when a rollout must keep critical tasks available when the model endpoint is impaired, but here a noninteractive document service runs overnight but consumes the premium synchronous request budget.

C:

Rare slow requests determine user-facing SLA risk that median timing can conceal; it is justified if a contact center needs 95th-percentile responses under its contractual target, not the case in which a noninteractive document service runs overnight but consumes the premium synchronous request budget.

D:

Accuracy improvement alone does not establish feasibility when the deployed workload violates unit economics; its scope includes cases where a sales assistant beats baseline accuracy but costs three times the approved cost per transaction, whereas the problem presented is that a noninteractive document service runs overnight but consumes the premium synchronous request budget.

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