Anthropic CCAR-P (Claude Certified Architect - Professional) Practice Test 22

Question 210

Six nearly identical versions of a policy dominate the top ten retrieved passages. What outcome must the reviewer see before proceeding?

  1. The measured evidence confirms that cross-tenant probes return no forbidden passages even when semantic similarity is high.
  2. The scenario test demonstrates that top-ranked results contain unique current policy passages rather than redundant versions.
  3. The end-to-end check confirms that recall at the passage level improves against an annotated answer-location benchmark.
  4. The observed outcome is that revoked passages disappear and new documents become searchable after the declared freshness window.

Explanation

Correct Answer

B:

Duplicate and stale passages waste context and can bias synthesis toward conflicting historic content; the test is necessary where search results show duplicate near-identical chunks from many versions of the same policy.

Incorrect Answers

A:

Vector similarity alone cannot enforce source entitlement or policy validity; this makes sense when a corpus contains both general guidance and document-specific restrictions that should not cross tenants, rather than the risk that six nearly identical versions of a policy dominate the top ten retrieved passages.

C:

Document-level relevance is insufficient when the model needs precise passages to ground claims; the trigger would be that an assistant retrieves relevant file names but not the answer-bearing paragraphs, while the immediate question involves six nearly identical versions of a policy dominate the top ten retrieved passages.

D:

Freshness requires consistent ingestion and invalidation rather than prompt reminders about current policy; the design helps when a knowledge base is refreshed nightly but answers still cite superseded policies, but the present constraint is that six nearly identical versions of a policy dominate the top ten retrieved passages.

Question 211

Users ask for exact invoice IDs that embeddings sometimes match to similar-looking IDs. What should be incorporated into the approved design?

  1. Within the rollout design, use semantic retrieval with reranking over task-relevant passages.
  2. For the next architecture revision, apply strict source filters and citation validation before response generation.
  3. For the current integration, use lexical or keyword matching for exact identifiers with typed filters.
  4. Under the system plan, use hybrid lexical and vector retrieval followed by relevance reranking.

Explanation

Correct Answer

C:

Exact-match retrieval is more reliable for literal codes than approximate semantic similarity; acceptance evidence must confirm that identifier probes retrieve precisely the requested invoice and not neighboring numbers.

Incorrect Answers

A:

Embedding-based similarity can connect paraphrased symptoms to documentation that shares meaning not wording; the other workload is one where a support engineer describes symptoms in different words from the troubleshooting manual, not the risk where users ask for exact invoice IDs that embeddings sometimes match to similar-looking IDs.

B:

Authority-constrained retrieval is essential when a semantically relevant unofficial page is inadmissible; this applies when a customer wants answers only from a small set of signed policy documents, not this situation where users ask for exact invoice IDs that embeddings sometimes match to similar-looking IDs.

D:

Hybrid search preserves exact statutory references while still matching paraphrased context; it is justified if compliance questions combine a statute number with a natural-language risk description, but here users ask for exact invoice IDs that embeddings sometimes match to similar-looking IDs.

Question 212

A support engineer describes symptoms in different words from the troubleshooting manual. Which approach should the design team choose?

  1. For the affected service, combine structured joins or graph traversal with retrieved textual evidence.
  2. In the project scope, query the authoritative transactional service at request time instead of relying on stale RAG.
  3. Within the rollout design, apply strict source filters and citation validation before response generation.
  4. For the next architecture revision, use semantic retrieval with reranking over task-relevant passages.

Explanation

Correct Answer

D:

Embedding-based similarity can connect paraphrased symptoms to documentation that shares meaning not wording; production testing should demonstrate that paraphrased benchmark queries retrieve relevant repair procedures despite vocabulary differences.

Incorrect Answers

A:

Relationships and entity constraints can be lost when treated as independent similarity matches; it serves conditions where a researcher asks for multi-hop relationships across linked supplier records, but this decision concerns a support engineer describes symptoms in different words from the troubleshooting manual.

B:

Fast-changing structured state requires live, authorized lookup rather than static document similarity; the other case is when an assistant answers 'What is the current account balance?' from a monthly indexed snapshot, but the actual risk is that a support engineer describes symptoms in different words from the troubleshooting manual.

C:

Authority-constrained retrieval is essential when a semantically relevant unofficial page is inadmissible; this makes sense when a customer wants answers only from a small set of signed policy documents, but the requested boundary concerns a support engineer describes symptoms in different words from the troubleshooting manual.

Question 213

A compliance assistant must answer queries containing exact statute identifiers and paraphrased risk descriptions, while citing only a small set of approved signed policies. Which TWO retrieval design choices jointly provide match quality and source admissibility? Choose TWO.

  1. For the primary workstream, use semantic retrieval with reranking over task-relevant passages.
  2. For this workflow, apply strict source filters and citation validation before response generation.
  3. Under the reviewed approach, combine structured joins or graph traversal with retrieved textual evidence.
  4. In the project scope, use hybrid lexical and vector retrieval followed by relevance reranking.
  5. Within the rollout design, use lexical or keyword matching for exact identifiers with typed filters.

Explanation

Correct Answers

B:

Authority-constrained retrieval is essential when a semantically relevant unofficial page is inadmissible; the controlled test needs to find that every generated claim is supported by a permitted signed source in the acceptance suite; another obligation arises when compliance questions combine a statute number with a natural-language risk description.

D:

Hybrid search preserves exact statutory references while still matching paraphrased context; the go-live check should establish that evaluation covers both citation-ID accuracy and semantic question coverage; another obligation arises when a customer wants answers only from a small set of signed policy documents.

Incorrect Answers

A:

Embedding-based similarity can connect paraphrased symptoms to documentation that shares meaning not wording; it serves a setting where a support engineer describes symptoms in different words from the troubleshooting manual, rather than the condition that a customer wants answers only from a small set of signed policy documents; another obligation arises when compliance questions combine a statute number with a natural-language risk description.

C:

Relationships and entity constraints can be lost when treated as independent similarity matches; the other workload is one where a researcher asks for multi-hop relationships across linked supplier records, but the present constraint is that a customer wants answers only from a small set of signed policy documents; another obligation arises when compliance questions combine a statute number with a natural-language risk description.

E:

Exact-match retrieval is more reliable for literal codes than approximate semantic similarity; the alternative helps if users ask for exact invoice IDs that embeddings sometimes match to similar-looking IDs, whereas the release criterion addresses a customer wants answers only from a small set of signed policy documents; another obligation arises when compliance questions combine a statute number with a natural-language risk description.

Question 214

An assistant answers 'What is the current account balance?' from a monthly indexed snapshot. What revision would make the proposal more defensible?

  1. Within the delivery plan, combine structured joins or graph traversal with retrieved textual evidence.
  2. At the system boundary, use hybrid lexical and vector retrieval followed by relevance reranking.
  3. In the initial solution, use semantic retrieval with reranking over task-relevant passages.
  4. For this workflow, query the authoritative transactional service at request time instead of relying on stale RAG.

Explanation

Correct Answer

D:

Fast-changing structured state requires live, authorized lookup rather than static document similarity; validation needs to show that balance responses match the ledger after a simulated transaction update.

Incorrect Answers

A:

Relationships and entity constraints can be lost when treated as independent similarity matches; that belongs to situations where a researcher asks for multi-hop relationships across linked supplier records, not this situation where an assistant answers 'What is the current account balance?' from a monthly indexed snapshot.

B:

Hybrid search preserves exact statutory references while still matching paraphrased context; it is relevant if compliance questions combine a statute number with a natural-language risk description, whereas this scenario involves an assistant answers 'What is the current account balance?' from a monthly indexed snapshot.

C:

Embedding-based similarity can connect paraphrased symptoms to documentation that shares meaning not wording; it serves conditions where a support engineer describes symptoms in different words from the troubleshooting manual, but here an assistant answers 'What is the current account balance?' from a monthly indexed snapshot.

Question 215

A researcher asks for multi-hop relationships across linked supplier records. How should the architect revise the proposed solution?

  1. For this production path, apply strict source filters and citation validation before response generation.
  2. For the intended users, combine structured joins or graph traversal with retrieved textual evidence.
  3. Within the delivery plan, query the authoritative transactional service at request time instead of relying on stale RAG.
  4. At the system boundary, use lexical or keyword matching for exact identifiers with typed filters.
  5. For the primary workstream, use hybrid lexical and vector retrieval followed by relevance reranking.

Explanation

Correct Answer

B:

Relationships and entity constraints can be lost when treated as independent similarity matches; a clear sign-off signal is that multi-hop test cases preserve entity identities across each relation and cited document.

Incorrect Answers

A:

Authority-constrained retrieval is essential when a semantically relevant unofficial page is inadmissible; this control is needed when a customer wants answers only from a small set of signed policy documents, but the actual risk is that a researcher asks for multi-hop relationships across linked supplier records.

C:

Fast-changing structured state requires live, authorized lookup rather than static document similarity; it serves a setting where an assistant answers 'What is the current account balance?' from a monthly indexed snapshot, not the present case where a researcher asks for multi-hop relationships across linked supplier records.

D:

Exact-match retrieval is more reliable for literal codes than approximate semantic similarity; that is defensible if users ask for exact invoice IDs that embeddings sometimes match to similar-looking IDs, while this problem involves a researcher asks for multi-hop relationships across linked supplier records.

E:

Hybrid search preserves exact statutory references while still matching paraphrased context; the other workload is one where compliance questions combine a statute number with a natural-language risk description, and not where a researcher asks for multi-hop relationships across linked supplier records.

Question 216

A supplier-risk research service may cite only signed policies, must follow indirect ownership links among suppliers, and must summarize portfolio-wide trends rather than a single paragraph. Which THREE retrieval or aggregation techniques are needed for those separate answer types and evidence boundaries? Choose THREE.

  1. In the proposed design, apply strict source filters and citation validation before response generation.
  2. Under the reviewed approach, use hybrid lexical and vector retrieval followed by relevance reranking.
  3. For this production path, combine structured joins or graph traversal with retrieved textual evidence.
  4. For the intended users, use lexical or keyword matching for exact identifiers with typed filters.
  5. Within the delivery plan, aggregate and sample relevant corpus segments before synthesis with coverage checks.
  6. At the system boundary, use semantic retrieval with reranking over task-relevant passages.

Explanation

Correct Answers

A:

Authority-constrained retrieval is essential when a semantically relevant unofficial page is inadmissible; the controlled test needs to find that every generated claim is supported by a permitted signed source in the acceptance suite; another obligation arises when a query asks for broad document trends rather than a single matching passage; the additional independent risk is that a researcher asks for multi-hop relationships across linked supplier records.

C:

Relationships and entity constraints can be lost when treated as independent similarity matches; the go-live check should establish that multi-hop test cases preserve entity identities across each relation and cited document; another obligation arises when a customer wants answers only from a small set of signed policy documents; the additional independent risk is that a query asks for broad document trends rather than a single matching passage.

E:

Top-k retrieval optimized for point answers can underrepresent the corpus in global trend analysis; the solution is testable when trend reports represent key categories across the full selected period rather than a few high-scoring chunks; another obligation arises when a customer wants answers only from a small set of signed policy documents; the additional independent risk is that a researcher asks for multi-hop relationships across linked supplier records.

Incorrect Answers

B:

Hybrid search preserves exact statutory references while still matching paraphrased context; the approach helps when compliance questions combine a statute number with a natural-language risk description, but the present constraint is that a query asks for broad document trends rather than a single matching passage; another obligation arises when a customer wants answers only from a small set of signed policy documents; the additional independent risk is that a researcher asks for multi-hop relationships across linked supplier records.

D:

Exact-match retrieval is more reliable for literal codes than approximate semantic similarity; it is relevant if users ask for exact invoice IDs that embeddings sometimes match to similar-looking IDs, whereas the release criterion addresses a customer wants answers only from a small set of signed policy documents; another obligation arises when a query asks for broad document trends rather than a single matching passage; the additional independent risk is that a researcher asks for multi-hop relationships across linked supplier records.

F:

Embedding-based similarity can connect paraphrased symptoms to documentation that shares meaning not wording; the other case is when a support engineer describes symptoms in different words from the troubleshooting manual, not the risk where a researcher asks for multi-hop relationships across linked supplier records; another obligation arises when a customer wants answers only from a small set of signed policy documents; the additional independent risk is that a query asks for broad document trends rather than a single matching passage.

Question 217

A query asks for broad document trends rather than a single matching passage. What intervention best matches the stated boundary?

  1. As a candidate control, aggregate and sample relevant corpus segments before synthesis with coverage checks.
  2. In the initial solution, query the authoritative transactional service at request time instead of relying on stale RAG.
  3. In the proposed design, use semantic retrieval with reranking over task-relevant passages.
  4. In the release design, combine structured joins or graph traversal with retrieved textual evidence.

Explanation

Correct Answer

A:

Top-k retrieval optimized for point answers can underrepresent the corpus in global trend analysis; acceptance evidence must confirm that trend reports represent key categories across the full selected period rather than a few high-scoring chunks.

Incorrect Answers

B:

Fast-changing structured state requires live, authorized lookup rather than static document similarity; that fits cases where an assistant answers 'What is the current account balance?' from a monthly indexed snapshot, but here a query asks for broad document trends rather than a single matching passage.

C:

Embedding-based similarity can connect paraphrased symptoms to documentation that shares meaning not wording; this control is needed when a support engineer describes symptoms in different words from the troubleshooting manual, instead of when a query asks for broad document trends rather than a single matching passage.

D:

Relationships and entity constraints can be lost when treated as independent similarity matches; that would be useful if a researcher asks for multi-hop relationships across linked supplier records, not the case in which a query asks for broad document trends rather than a single matching passage.

Question 218

A customer types a precise purchase order number and search returns a semantically similar, wrong identifier. What evidence would reveal whether the control truly works?

  1. The controlled replay verifies that trend reports represent key categories across the full selected period rather than a few high-scoring chunks.
  2. A reproducible run shows that identifier probes retrieve precisely the requested invoice and not neighboring numbers.
  3. The signed test record confirms that paraphrased benchmark queries retrieve relevant repair procedures despite vocabulary differences.
  4. The sign-off record demonstrates that evaluation covers both citation-ID accuracy and semantic question coverage.

Explanation

Correct Answer

B:

Exact-match retrieval is more reliable for literal codes than approximate semantic similarity; this is relevant for cases where users ask for exact invoice IDs that embeddings sometimes match to similar-looking IDs.

Incorrect Answers

A:

Top-k retrieval optimized for point answers can underrepresent the corpus in global trend analysis; it is justified if a query asks for broad document trends rather than a single matching passage, but the requested boundary concerns a customer types a precise purchase order number and search returns a semantically similar, wrong identifier.

C:

Embedding-based similarity can connect paraphrased symptoms to documentation that shares meaning not wording; it targets cases where a support engineer describes symptoms in different words from the troubleshooting manual, while this problem involves a customer types a precise purchase order number and search returns a semantically similar, wrong identifier.

D:

Hybrid search preserves exact statutory references while still matching paraphrased context; the approach helps when compliance questions combine a statute number with a natural-language risk description, and not where a customer types a precise purchase order number and search returns a semantically similar, wrong identifier.

Question 219

An engineer describes an outage symptom colloquially and search misses the manual's technical wording. Which result proves the requested control, not just configuration?

  1. The reviewer observes that every generated claim is supported by a permitted signed source in the acceptance suite.
  2. The release report verifies that evaluation covers both citation-ID accuracy and semantic question coverage.
  3. The controlled replay verifies that multi-hop test cases preserve entity identities across each relation and cited document.
  4. A reproducible run shows that paraphrased benchmark queries retrieve relevant repair procedures despite vocabulary differences.

Explanation

Correct Answer

D:

Embedding-based similarity can connect paraphrased symptoms to documentation that shares meaning not wording; this is a functional check for the case where a support engineer describes symptoms in different words from the troubleshooting manual.

Incorrect Answers

A:

Authority-constrained retrieval is essential when a semantically relevant unofficial page is inadmissible; this technique fits where a customer wants answers only from a small set of signed policy documents, but the present constraint is that an engineer describes an outage symptom colloquially and search misses the manual's technical wording.

B:

Hybrid search preserves exact statutory references while still matching paraphrased context; the trigger would be that compliance questions combine a statute number with a natural-language risk description, whereas the current issue is that an engineer describes an outage symptom colloquially and search misses the manual's technical wording.

C:

Relationships and entity constraints can be lost when treated as independent similarity matches; the design helps when a researcher asks for multi-hop relationships across linked supplier records, whereas the release criterion addresses an engineer describes an outage symptom colloquially and search misses the manual's technical wording.

Question 220

A legal query has both a citation code and a prose concern, but search matches only one component. Which acceptance result provides the most specific evidence?

  1. A production-like test shows that multi-hop test cases preserve entity identities across each relation and cited document.
  2. The release check confirms that balance responses match the ledger after a simulated transaction update.
  3. The reviewer observes that paraphrased benchmark queries retrieve relevant repair procedures despite vocabulary differences.
  4. The scenario test demonstrates that evaluation covers both citation-ID accuracy and semantic question coverage.

Explanation

Correct Answer

D:

Hybrid search preserves exact statutory references while still matching paraphrased context; this checks the situation in which compliance questions combine a statute number with a natural-language risk description.

Incorrect Answers

A:

Relationships and entity constraints can be lost when treated as independent similarity matches; this applies when a researcher asks for multi-hop relationships across linked supplier records, but here a legal query has both a citation code and a prose concern, but search matches only one component.

B:

Fast-changing structured state requires live, authorized lookup rather than static document similarity; the alternative helps if an assistant answers 'What is the current account balance?' from a monthly indexed snapshot, instead of when a legal query has both a citation code and a prose concern, but search matches only one component.

C:

Embedding-based similarity can connect paraphrased symptoms to documentation that shares meaning not wording; it is justified if a support engineer describes symptoms in different words from the troubleshooting manual, not the case in which a legal query has both a citation code and a prose concern, but search matches only one component.

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