Question 290
An architecture board is assessing 2 separate deliverables: (1) avoid long generated replies when a client requires a short summary or fixed output budget; and (2) keep permanent application behavior separate from ordinary user requests. Which distinct implementations meet these deliverables? Choose TWO.
- Constrain response length for concise summaries
- Use explicit system and developer instructions for policy
- Lower sampling temperature for stable classifications
- Use explicit few-shot examples for format nuance
- Batch independent embeddings for ingestion efficiency
- Cache retrieved context with source versioning
Explanation
Correct Answers
A:For the stated requirement to avoid long generated replies when a client requires a short summary or fixed output budget, a sensible output-token cap and concise instruction bound generation size, but a cap that is too small may truncate content; this is the mechanism that the system must implement. This choice satisfies the specific deliverable to avoid long generated replies when a client requires a short summary or fixed output budget, while the other deliverables require separate capabilities.
B:For the stated requirement to keep permanent application behavior separate from ordinary user requests, high-priority instructions establish the app's role and safety boundaries, but external authorization still requires programmatic enforcement; this is the mechanism that the system must implement. This choice satisfies the specific deliverable to keep permanent application behavior separate from ordinary user requests, while the other deliverables require separate capabilities.
Incorrect Answers
C:For a system whose objective is to reduce unwarranted variability when an output must follow a fixed decision rubric, lower temperature can reduce sampling randomness where supported; it does not guarantee factual accuracy or replace schema validation; this does not directly meet the required outcome to keep permanent application behavior separate from ordinary user requests.
D:For a system whose objective is to teach a model a specialized labeling convention that general instructions describe poorly, well-chosen examples illustrate the desired input-output mapping, including ambiguous edge cases; they must be representative; this does not directly meet the required outcome to keep permanent application behavior separate from ordinary user requests.
E:For a system whose objective is to lower overhead when indexing many small documents without changing semantic meaning, batching compatible embedding requests can improve throughput within service rate limits; errors require granular retries; this does not directly meet the required outcome to keep permanent application behavior separate from ordinary user requests.
F:For a system whose objective is to avoid repeated retrieval for unchanged documents while preventing stale grounding, versioned cache keys or invalidation preserve freshness when approved source content changes; this does not directly meet the required outcome to keep permanent application behavior separate from ordinary user requests.
Question 291
An Azure AI workload passes basic health probes but has this issue: An ingestion job sends thousands of tiny embedding calls serially and cannot finish its nightly window. Which fix targets it?
- Assess false positives with a labeled set and tune supported filter settings under policy approval
- Classify task complexity and route easy cases to a validated lower-cost model
- Implement a supported no-evidence response and escalation route for low-support cases
- Group independent embedding requests into supported batches with rate-aware retries
Explanation
Correct Answer
D:The fault signal is that an ingestion job sends thousands of tiny embedding calls serially and cannot finish its nightly window; the targeted fix works because batching compatible embedding requests can improve throughput within service rate limits; errors require granular retries, restoring the ability to lower overhead when indexing many small documents without changing semantic meaning.
Incorrect Answers
A:If the failing function were to reduce legitimate request blocks without weakening essential guardrails, this change could be relevant because precision and recall of safety classifications require policy-labeled evaluation rather than a blanket filter disablement; the actual symptom is that an ingestion job sends thousands of tiny embedding calls serially and cannot finish its nightly window, which calls for a different repair.
B:If the failing function were to send bounded simple queries to economical models and harder reasoning tasks to stronger ones, this change could be relevant because a routing policy can select among models based on evaluated task difficulty and quality constraints; the actual symptom is that an ingestion job sends thousands of tiny embedding calls serially and cannot finish its nightly window, which calls for a different repair.
C:If the failing function were to prevent confident unsupported answers when a retrieval score or evidence threshold is weak, this change could be relevant because a calibrated no-answer path is safer than fabricating a missing fact; thresholds should be tuned on labeled questions; the actual symptom is that an ingestion job sends thousands of tiny embedding calls serially and cannot finish its nightly window, which calls for a different repair.
Question 292
A development group must improve an initial generated report by checking it against a defined factual and style rubric. What is the most appropriate Azure AI implementation?
- Inspect prompt regressions using paired examples
- Add a critique-and-revise stage for draft quality
- Use explicit few-shot examples for format nuance
- Use external evidence to verify generated claims
Explanation
Correct Answer
B:For the stated requirement to improve an initial generated report by checking it against a defined factual and style rubric, a critique step can identify defects in an earlier draft when grounded in explicit criteria; self-review alone is not an independent fact source; this is the mechanism that the system must implement.
Incorrect Answers
A:For a system whose objective is to detect whether changing instructions improves one use case while damaging another, paired cases give controlled evidence about prompt effects and expose class-specific regressions; this does not directly meet the required outcome to improve an initial generated report by checking it against a defined factual and style rubric.
C:For a system whose objective is to teach a model a specialized labeling convention that general instructions describe poorly, well-chosen examples illustrate the desired input-output mapping, including ambiguous edge cases; they must be representative; this does not directly meet the required outcome to improve an initial generated report by checking it against a defined factual and style rubric.
D:For a system whose objective is to avoid treating a model's agreement with its own answer as proof of correctness, independent source retrieval or deterministic tools can verify claims beyond the generator's self-confidence; this does not directly meet the required outcome to improve an initial generated report by checking it against a defined factual and style rubric.
Question 293
After implementing a workflow to inspect prompt regressions using paired examples, the team needs a robust release gate. Which check should it choose?
- Measure rubric violations and unsupported statements before and after revision
- Report per-category deltas and inspect failed examples instead of relying on one aggregate score
- Evaluate task accuracy, p95 latency, and cost per successful output
- Measure indexed chunks per minute, per-item failures, and sustained quota usage
Explanation
Correct Answer
B:The accepted change is to inspect prompt regressions using paired examples; this evaluation checks whether it can detect whether changing instructions improves one use case while damaging another, since paired cases give controlled evidence about prompt effects and expose class-specific regressions.
Incorrect Answers
A:This test concerns the ability to improve an initial generated report by checking it against a defined factual and style rubric, for which a critique step can identify defects in an earlier draft when grounded in explicit criteria; self-review alone is not an independent fact source; it does not establish whether the change to inspect prompt regressions using paired examples works for the current release.
C:This test concerns the ability to choose between smaller and larger models based on business quality and cost, for which representative evaluations reveal the trade-off among quality, latency and spend; model size is an imperfect proxy; it does not establish whether the change to inspect prompt regressions using paired examples works for the current release.
D:This test concerns the ability to lower overhead when indexing many small documents without changing semantic meaning, for which batching compatible embedding requests can improve throughput within service rate limits; errors require granular retries; it does not establish whether the change to inspect prompt regressions using paired examples works for the current release.
Question 294
Production telemetry reveals this failure: Two model passes repeat the same invented regulation and incorrectly mark it verified. Which change most directly corrects it?
- Cross-check each material claim against approved authoritative sources or executable rules
- Specify an appropriate concise response format and cap generation output within the supported model settings
- Run the same labeled task set across candidate models and compare operational metrics
- Compare fixed regression sets covering both summarization and citation tasks across prompt versions
Explanation
Correct Answer
A:The fault signal is that two model passes repeat the same invented regulation and incorrectly mark it verified; the targeted fix works because independent source retrieval or deterministic tools can verify claims beyond the generator's self-confidence, restoring the ability to avoid treating a model's agreement with its own answer as proof of correctness.
Incorrect Answers
B:If the failing function were to avoid long generated replies when a client requires a short summary or fixed output budget, this change could be relevant because a sensible output-token cap and concise instruction bound generation size, but a cap that is too small may truncate content; the actual symptom is that two model passes repeat the same invented regulation and incorrectly mark it verified, which calls for a different repair.
C:If the failing function were to choose between smaller and larger models based on business quality and cost, this change could be relevant because representative evaluations reveal the trade-off among quality, latency and spend; model size is an imperfect proxy; the actual symptom is that two model passes repeat the same invented regulation and incorrectly mark it verified, which calls for a different repair.
D:If the failing function were to detect whether changing instructions improves one use case while damaging another, this change could be relevant because paired cases give controlled evidence about prompt effects and expose class-specific regressions; the actual symptom is that two model passes repeat the same invented regulation and incorrectly mark it verified, which calls for a different repair.
Question 295
A system now includes a component to cache retrieved context with source versioning. Which evaluation should accompany the rollout?
- Measure context-token reduction and preservation of active task facts
- Check stale-citation incidence and cache hits after source updates
- Compare p95 latency by span and count budget overruns per stage
- Inject rule-violating JSON fixtures and verify every violation is caught
Explanation
Correct Answer
B:The accepted change is to cache retrieved context with source versioning; this evaluation checks whether it can avoid repeated retrieval for unchanged documents while preventing stale grounding, since versioned cache keys or invalidation preserve freshness when approved source content changes.
Incorrect Answers
A:This test concerns the ability to reduce token cost in long sessions while keeping current commitments and unresolved tasks, for which selective compression retains required state but excludes repeated old messages; simply raising token limits does not fix unnecessary spend; it does not establish whether the change to cache retrieved context with source versioning works for the current release.
C:This test concerns the ability to identify which dependent stage to optimize in a slow multi-step AI workflow, for which stage-specific latency budgets reveal the true critical path across inference, retrieval and tools; it does not establish whether the change to cache retrieved context with source versioning works for the current release.
D:This test concerns the ability to catch schema-valid yet impossible business outputs before a workflow proceeds, for which a type-correct JSON value may still violate business invariants; deterministic post-validation enforces the rules; it does not establish whether the change to cache retrieved context with source versioning works for the current release.
Question 296
The support ticket describes this behavior: A team chooses the most expensive model without testing whether a smaller one meets its labeling requirements. Which engineering change should come first?
- Run the same labeled task set across candidate models and compare operational metrics
- Evaluate the draft against a rubric and regenerate only the failing sections using source evidence
- Compare fixed regression sets covering both summarization and citation tasks across prompt versions
- Run a business-rule validator after schema checks and reject impossible state transitions
Explanation
Correct Answer
A:The fault signal is that a team chooses the most expensive model without testing whether a smaller one meets its labeling requirements; the targeted fix works because representative evaluations reveal the trade-off among quality, latency and spend; model size is an imperfect proxy, restoring the ability to choose between smaller and larger models based on business quality and cost.
Incorrect Answers
B:If the failing function were to improve an initial generated report by checking it against a defined factual and style rubric, this change could be relevant because a critique step can identify defects in an earlier draft when grounded in explicit criteria; self-review alone is not an independent fact source; the actual symptom is that a team chooses the most expensive model without testing whether a smaller one meets its labeling requirements, which calls for a different repair.
C:If the failing function were to detect whether changing instructions improves one use case while damaging another, this change could be relevant because paired cases give controlled evidence about prompt effects and expose class-specific regressions; the actual symptom is that a team chooses the most expensive model without testing whether a smaller one meets its labeling requirements, which calls for a different repair.
D:If the failing function were to catch schema-valid yet impossible business outputs before a workflow proceeds, this change could be relevant because a type-correct JSON value may still violate business invariants; deterministic post-validation enforces the rules; the actual symptom is that a team chooses the most expensive model without testing whether a smaller one meets its labeling requirements, which calls for a different repair.
Question 297
After implementing a workflow to compare model candidates with task-specific benchmarks, the team needs a robust release gate. Which check should it choose?
- Test conflicting user requests for policy adherence and unauthorized tool execution
- Report per-category deltas and inspect failed examples instead of relying on one aggregate score
- Check stale-citation incidence and cache hits after source updates
- Measure claims corroborated by independent evidence and unresolved discrepancies
- Evaluate task accuracy, p95 latency, and cost per successful output
Explanation
Correct Answer
E:The accepted change is to compare model candidates with task-specific benchmarks; this evaluation checks whether it can choose between smaller and larger models based on business quality and cost, since representative evaluations reveal the trade-off among quality, latency and spend; model size is an imperfect proxy.
Incorrect Answers
A:This test concerns the ability to keep permanent application behavior separate from ordinary user requests, for which high-priority instructions establish the app's role and safety boundaries, but external authorization still requires programmatic enforcement; it does not establish whether the change to compare model candidates with task-specific benchmarks works for the current release.
B:This test concerns the ability to detect whether changing instructions improves one use case while damaging another, for which paired cases give controlled evidence about prompt effects and expose class-specific regressions; it does not establish whether the change to compare model candidates with task-specific benchmarks works for the current release.
C:This test concerns the ability to avoid repeated retrieval for unchanged documents while preventing stale grounding, for which versioned cache keys or invalidation preserve freshness when approved source content changes; it does not establish whether the change to compare model candidates with task-specific benchmarks works for the current release.
D:This test concerns the ability to avoid treating a model's agreement with its own answer as proof of correctness, for which independent source retrieval or deterministic tools can verify claims beyond the generator's self-confidence; it does not establish whether the change to compare model candidates with task-specific benchmarks works for the current release.
Question 298
A project needs to complete all of these distinct tasks: (1) catch schema-valid yet impossible business outputs before a workflow proceeds; and (2) send bounded simple queries to economical models and harder reasoning tasks to stronger ones. Select one suitable capability for each task. Choose TWO.
- Use explicit system and developer instructions for policy
- Cache retrieved context with source versioning
- Set latency budgets for model and tool spans
- Apply output validation with deterministic rules
- Use external evidence to verify generated claims
- Use model routing by task complexity
Explanation
Correct Answers
D:For the stated requirement to catch schema-valid yet impossible business outputs before a workflow proceeds, a type-correct JSON value may still violate business invariants; deterministic post-validation enforces the rules; this is the mechanism that the system must implement. This choice satisfies the specific deliverable to catch schema-valid yet impossible business outputs before a workflow proceeds, while the other deliverables require separate capabilities.
F:For the stated requirement to send bounded simple queries to economical models and harder reasoning tasks to stronger ones, a routing policy can select among models based on evaluated task difficulty and quality constraints; this is the mechanism that the system must implement. This choice satisfies the specific deliverable to send bounded simple queries to economical models and harder reasoning tasks to stronger ones, while the other deliverables require separate capabilities.
Incorrect Answers
A:For a system whose objective is to keep permanent application behavior separate from ordinary user requests, high-priority instructions establish the app's role and safety boundaries, but external authorization still requires programmatic enforcement; this does not directly meet the required outcome to catch schema-valid yet impossible business outputs before a workflow proceeds.
B:For a system whose objective is to avoid repeated retrieval for unchanged documents while preventing stale grounding, versioned cache keys or invalidation preserve freshness when approved source content changes; this does not directly meet the required outcome to catch schema-valid yet impossible business outputs before a workflow proceeds.
C:For a system whose objective is to identify which dependent stage to optimize in a slow multi-step AI workflow, stage-specific latency budgets reveal the true critical path across inference, retrieval and tools; this does not directly meet the required outcome to catch schema-valid yet impossible business outputs before a workflow proceeds.
E:For a system whose objective is to avoid treating a model's agreement with its own answer as proof of correctness, independent source retrieval or deterministic tools can verify claims beyond the generator's self-confidence; this does not directly meet the required outcome to catch schema-valid yet impossible business outputs before a workflow proceeds.
Question 299
The acceptance criteria for a production AI solution require it to choose between smaller and larger models based on business quality and cost. Which design should be approved?
- Compare model candidates with task-specific benchmarks
- Apply output validation with deterministic rules
- Use external evidence to verify generated claims
- Add a critique-and-revise stage for draft quality
Explanation
Correct Answer
A:For the stated requirement to choose between smaller and larger models based on business quality and cost, representative evaluations reveal the trade-off among quality, latency and spend; model size is an imperfect proxy; this is the mechanism that the system must implement.
Incorrect Answers
B:For a system whose objective is to catch schema-valid yet impossible business outputs before a workflow proceeds, a type-correct JSON value may still violate business invariants; deterministic post-validation enforces the rules; this does not directly meet the required outcome to choose between smaller and larger models based on business quality and cost.
C:For a system whose objective is to avoid treating a model's agreement with its own answer as proof of correctness, independent source retrieval or deterministic tools can verify claims beyond the generator's self-confidence; this does not directly meet the required outcome to choose between smaller and larger models based on business quality and cost.
D:For a system whose objective is to improve an initial generated report by checking it against a defined factual and style rubric, a critique step can identify defects in an earlier draft when grounded in explicit criteria; self-review alone is not an independent fact source; this does not directly meet the required outcome to choose between smaller and larger models based on business quality and cost.
Question 300
During a regression test, the following unexpected result occurs: One-paragraph summaries frequently exceed the service's character budget and delay clients. What should the engineer do?
- Specify an appropriate concise response format and cap generation output within the supported model settings
- Reduce stochastic sampling and validate the mapping on labeled repeated tests
- Cross-check each material claim against approved authoritative sources or executable rules
- Add a small diverse set of labeled examples with boundary cases to the prompt
Explanation
Correct Answer
A:The fault signal is that one-paragraph summaries frequently exceed the service's character budget and delay clients; the targeted fix works because a sensible output-token cap and concise instruction bound generation size, but a cap that is too small may truncate content, restoring the ability to avoid long generated replies when a client requires a short summary or fixed output budget.
Incorrect Answers
B:If the failing function were to reduce unwarranted variability when an output must follow a fixed decision rubric, this change could be relevant because lower temperature can reduce sampling randomness where supported; it does not guarantee factual accuracy or replace schema validation; the actual symptom is that one-paragraph summaries frequently exceed the service's character budget and delay clients, which calls for a different repair.
C:If the failing function were to avoid treating a model's agreement with its own answer as proof of correctness, this change could be relevant because independent source retrieval or deterministic tools can verify claims beyond the generator's self-confidence; the actual symptom is that one-paragraph summaries frequently exceed the service's character budget and delay clients, which calls for a different repair.
D:If the failing function were to teach a model a specialized labeling convention that general instructions describe poorly, this change could be relevant because well-chosen examples illustrate the desired input-output mapping, including ambiguous edge cases; they must be representative; the actual symptom is that one-paragraph summaries frequently exceed the service's character budget and delay clients, which calls for a different repair.