cert
cert-1
cert-2

Pass Google Professional Machine Learning Engineer Exam in First Attempt Guaranteed!

Get 100% Latest Exam Questions, Accurate & Verified Answers to Pass the Actual Exam!
30 Days Free Updates, Instant Download!

cert-5
cert-6
Professional Machine Learning Engineer Exam - Verified By Experts
Professional Machine Learning Engineer Premium Bundle
$29.98

Professional Machine Learning Engineer Premium Bundle

$79.99
$109.97
  • Premium File 402 Questions & Answers. Last update: Nov 13, 2026
  • Training Course 69 Video Lectures
  • Study Guide 376 Pages
 
$109.97
$79.99
accept 47 downloads in last 7 days
block-screenshots
Purchase Google Professional Machine Learning Engineer Exam Training Products Individually
Google Professional Machine Learning Engineer Premium File
Premium File 402 Q&A
$76.99$69.99
Google Professional Machine Learning Engineer Video Training Course
Training Course 69 Lectures
$21.99 $19.99
Google Professional Machine Learning Engineer PDF Study Guide
Study Guide 376 Pages
$21.99 $19.99
Google Professional Machine Learning Engineer Practice Test 82

Question 810

During a controlled replay of the data and model handoff to a second engineering team, the following condition occurs: A team imports a pretrained JAX checkpoint in a notebook but assumes its model weights and device setup will load unchanged in a PyTorch-only runtime. The handoff must preserve privacy, lineage and testable behavior. Which TWO selections form an appropriate before-and-after verification pair? Select TWO.

  1. A clean notebook loads the exact model revision, runs representative inputs and reports consistent outputs.
  2. The notebook can access allowed buckets without a stored private key and the old key no longer authenticates.
  3. The equivalent daily query scans only the eligible partitions and returns the same number of matched entities.
  4. Inspect query result sizes, runtime memory, pandas allocations and which aggregations can run in BigQuery SQL.
  5. Inspect saved artifact format, framework versions, device placement requirements and the model card's supported execution paths.

Explanation

Correct Answers

A:

A clean notebook loads the exact model revision, runs representative inputs and reports consistent outputs. This is specific acceptance evidence for model garden jax accelerator compatibility. Framework-native checkpoints and runtime APIs are not automatically interchangeable across ML libraries. It confirms the controlled correction: Choose a compatible Model Garden model/runtime or perform a validated checkpoint conversion before training.

E:

Inspect saved artifact format, framework versions, device placement requirements and the model card's supported execution paths. This inspects evidence relevant to model garden jax accelerator compatibility. Framework-native checkpoints and runtime APIs are not automatically interchangeable across ML libraries. After identifying the cause, the matching correction is: Choose a compatible Model Garden model/runtime or perform a validated checkpoint conversion before training.

Incorrect Answers

B:

The notebook can access allowed buckets without a stored private key and the old key no longer authenticates. This relates to workbench credentials embedded in repository but does not establish the requested before-and-after evidence. The actual diagnostic is: Inspect saved artifact format, framework versions, device placement requirements and the model card's supported execution paths. The specific acceptance result is: A clean notebook loads the exact model revision, runs representative inputs and reports consistent outputs.

C:

The equivalent daily query scans only the eligible partitions and returns the same number of matched entities. This relates to bigquery partition filter cost optimization but does not establish the requested before-and-after evidence. The actual diagnostic is: Inspect saved artifact format, framework versions, device placement requirements and the model card's supported execution paths. The specific acceptance result is: A clean notebook loads the exact model revision, runs representative inputs and reports consistent outputs.

D:

Inspect query result sizes, runtime memory, pandas allocations and which aggregations can run in BigQuery SQL. This relates to notebook memory exhaustion on full data extraction but does not establish the requested before-and-after evidence. The actual diagnostic is: Inspect saved artifact format, framework versions, device placement requirements and the model card's supported execution paths. The specific acceptance result is: A clean notebook loads the exact model revision, runs representative inputs and reports consistent outputs.

Question 811

During a controlled replay of the data and model handoff to a second engineering team, the following condition occurs: A distributed preprocessing job joins two huge feature tables, but one NULL-like customer category contains most rows and one reducer becomes the bottleneck. The handoff must preserve privacy, lineage and testable behavior. Select TWO: the most relevant root-cause check and the specific observation that proves the fix worked.

  1. Long-tail shuffle tasks shorten and the normalized output row count matches a reference implementation.
  2. Independent notebook runs resolve the same approved libraries and produce comparable metric and artifact outputs.
  3. Annotators can complete validated labels without seeing identifying data and the project retains compliant traceability.
  4. Inspect join-key frequency distributions, per-task runtime, shuffle spill and the output cardinality before optimization.
  5. Examine the partitioning column, query plan, bytes scanned and predicate form against the table's partition definition.

Explanation

Correct Answers

A:

Long-tail shuffle tasks shorten and the normalized output row count matches a reference implementation. This is specific acceptance evidence for managed spark join-key skew repair. Skew concentrates otherwise distributed work on a single hot key, which extra workers alone may not resolve. It confirms the controlled correction: Handle the dominant key separately or repartition and salt the skewed join while validating join semantics and counts.

D:

Inspect join-key frequency distributions, per-task runtime, shuffle spill and the output cardinality before optimization. This inspects evidence relevant to managed spark join-key skew repair. Skew concentrates otherwise distributed work on a single hot key, which extra workers alone may not resolve. After identifying the cause, the matching correction is: Handle the dominant key separately or repartition and salt the skewed join while validating join semantics and counts.

Incorrect Answers

B:

Independent notebook runs resolve the same approved libraries and produce comparable metric and artifact outputs. This relates to notebook dependency lock reproducibility but does not establish the requested before-and-after evidence. The actual diagnostic is: Inspect join-key frequency distributions, per-task runtime, shuffle spill and the output cardinality before optimization. The specific acceptance result is: Long-tail shuffle tasks shorten and the normalized output row count matches a reference implementation.

C:

Annotators can complete validated labels without seeing identifying data and the project retains compliant traceability. This relates to data minimization for labeling inputs but does not establish the requested before-and-after evidence. The actual diagnostic is: Inspect join-key frequency distributions, per-task runtime, shuffle spill and the output cardinality before optimization. The specific acceptance result is: Long-tail shuffle tasks shorten and the normalized output row count matches a reference implementation.

E:

Examine the partitioning column, query plan, bytes scanned and predicate form against the table's partition definition. This relates to bigquery partition filter cost optimization but does not establish the requested before-and-after evidence. The actual diagnostic is: Inspect join-key frequency distributions, per-task runtime, shuffle spill and the output cardinality before optimization. The specific acceptance result is: Long-tail shuffle tasks shorten and the normalized output row count matches a reference implementation.

Question 812

During a regulated MLOps pipeline deployment, the following condition occurs: A model registry entry lacks the upstream version of the training feature table, preventing investigators from reconstructing a reported prediction. The workflow must be repeatable and protect the last approved production model. Which technical explanation best accounts for the reported behavior?

  1. Orchestration without failure notification can leave models stale even when the last successful deployment was correct.
  2. Orchestrators often retry tasks; duplicate side effects are prevented by idempotent data operations.
  3. Model weights alone do not capture the data lineage necessary to reproduce and investigate predictions.
  4. Distributed in-memory object stores are not durable archival storage; failure recovery needs explicit design.

Explanation

Correct Answer

C:

Model weights alone do not capture the data lineage necessary to reproduce and investigate predictions. This explains the mechanism behind model feature lineage for audit and why the described failure is possible. A discriminating investigation is: Inspect ML Metadata execution links, table snapshots, transformations and model version lineage. The corresponding repair is: Record governed dataset snapshot and transformation versions as pipeline inputs linked to the resulting model artifact. This matters here because a model registry entry lacks the upstream version of the training feature table, preventing investigators from….

Incorrect Answers

A:

Orchestration without failure notification can leave models stale even when the last successful deployment was correct. This principle describes pipeline notification alert on silent failure, which is not the cause of the observed model feature lineage for audit failure. The controlling explanation is: Model weights alone do not capture the data lineage necessary to reproduce and investigate predictions.

B:

Orchestrators often retry tasks; duplicate side effects are prevented by idempotent data operations. This principle describes airflow task idempotency on retry, which is not the cause of the observed model feature lineage for audit failure. The controlling explanation is: Model weights alone do not capture the data lineage necessary to reproduce and investigate predictions.

D:

Distributed in-memory object stores are not durable archival storage; failure recovery needs explicit design. This principle describes ray distributed task failure isolation, which is not the cause of the observed model feature lineage for audit failure. The controlling explanation is: Model weights alone do not capture the data lineage necessary to reproduce and investigate predictions.

Question 813

While reproducing a training orchestration incident in isolated staging, the following condition occurs: A Kubeflow preprocessing task is evicted repeatedly because it requests little memory despite materializing large embeddings inside the container. The team requires an idempotent pipeline behavior and traceable evidence. What implementation change resolves the specific issue without substituting an unrelated fix?

  1. Package the fitted preprocessing component with the model and apply the same validated transform in both paths.
  2. Trigger candidate retraining only when justified evidence and sufficient labeled data are available, with evaluation gates.
  3. Tune component resource requests/limits and stream or partition embeddings rather than materializing every vector at once.
  4. Record governed dataset snapshot and transformation versions as pipeline inputs linked to the resulting model artifact.

Explanation

Correct Answer

C:

Tune component resource requests/limits and stream or partition embeddings rather than materializing every vector at once. This is the targeted corrective response for custom kfp component resource limits. Kubernetes resource limits constrain component runtime even when the model's training job itself has spare capacity. Verify the diagnosis using: Inspect Kubernetes task events, memory usage, component resource requests and container data-loading behavior. Confirm success using: The task completes within approved resources with predictable cost and produces the same number of embeddings. This matters here because while reproducing a training orchestration incident in isolated staging, the following condition occurs: A Kubeflow….

Incorrect Answers

A:

Package the fitted preprocessing component with the model and apply the same validated transform in both paths. This action addresses training serving preprocessing artifact sharing, not the failure in the described custom kfp component resource limits case. The targeted correction is: Tune component resource requests/limits and stream or partition embeddings rather than materializing every vector at once.

B:

Trigger candidate retraining only when justified evidence and sufficient labeled data are available, with evaluation gates. This action addresses data drift triggered retraining policy, not the failure in the described custom kfp component resource limits case. The targeted correction is: Tune component resource requests/limits and stream or partition embeddings rather than materializing every vector at once.

D:

Record governed dataset snapshot and transformation versions as pipeline inputs linked to the resulting model artifact. This action addresses model feature lineage for audit, not the failure in the described custom kfp component resource limits case. The targeted correction is: Tune component resource requests/limits and stream or partition embeddings rather than materializing every vector at once.

Question 814

While preparing a production training workflow on Gemini Enterprise Agent Platform, the following condition occurs: A transformer cannot fit its weights and activations on one device even at minimal batch size, but the team proposes more data replicas only. Engineers need an auditable, resource-aware correction to the issue. Which diagnostic check should be performed FIRST to isolate the described cause?

  1. Inspect region-zone capacity, quota limits, accelerator requirements and scheduling error events.
  2. Inspect feature computation cutoffs, window frame bounds, label timestamps and time-aware evaluation splits.
  3. Inspect model framework, accelerator kernel support, batch sizes, input pipeline speed and per-step cost.
  4. Inspect parameter and activation memory per device, optimizer memory and available interconnect bandwidth.

Explanation

Correct Answer

D:

Inspect parameter and activation memory per device, optimizer memory and available interconnect bandwidth. This inspects evidence relevant to model parallelism for memory-bound transformer. Data parallel replicas each hold a model copy; they do not solve a model-too-large-for-one-device problem. After identifying the cause, the matching correction is: Evaluate supported tensor or pipeline model parallelism to split the model across compatible accelerators. This matters here because while preparing a production training workflow on Gemini Enterprise Agent Platform, the following condition occurs:….

Incorrect Answers

A:

Inspect region-zone capacity, quota limits, accelerator requirements and scheduling error events. This examines accelerator quota and regional availability rather than the specific mechanism reported here. It is not the decisive first check for model parallelism for memory-bound transformer. The relevant evidence is: Inspect parameter and activation memory per device, optimizer memory and available interconnect bandwidth.

B:

Inspect feature computation cutoffs, window frame bounds, label timestamps and time-aware evaluation splits. This examines temporal training label leakage in rolling windows rather than the specific mechanism reported here. It is not the decisive first check for model parallelism for memory-bound transformer. The relevant evidence is: Inspect parameter and activation memory per device, optimizer memory and available interconnect bandwidth.

C:

Inspect model framework, accelerator kernel support, batch sizes, input pipeline speed and per-step cost. This examines tpu versus gpu for tensorflow workload rather than the specific mechanism reported here. It is not the decisive first check for model parallelism for memory-bound transformer. The relevant evidence is: Inspect parameter and activation memory per device, optimizer memory and available interconnect bandwidth. This matters here because while preparing a production training workflow on Gemini Enterprise Agent Platform, the following condition occurs:….

Question 815

During model-training validation on a Google Cloud experimental project, the following condition occurs: An operations team uses a general-purpose chat model to forecast thousands of numeric hourly warehouse counts but ignores established daily and weekly seasonality. The change should preserve a meaningful independent model-quality comparison. Which diagnostic check should be performed FIRST to isolate the described cause?

  1. Inspect task requirements for retrieval ranking, available query-document pairs, embedding similarity and labeled relevance examples.
  2. Determine whether failures arise from changing reference knowledge, stable behavior patterns or inadequate retrieval coverage.
  3. Inspect time stamps, autocorrelation, horizon, seasonal cycles and out-of-sample forecast metrics against traditional baselines.
  4. Compare pretraining benchmark scores, domain validation curves, trainable parameters and fine-tuning learning-rate choices.

Explanation

Correct Answer

C:

Inspect time stamps, autocorrelation, horizon, seasonal cycles and out-of-sample forecast metrics against traditional baselines. This inspects evidence relevant to forecast trend and seasonality versus llm. Structured numeric forecasting relies on temporal evidence; a conversational model is not automatically the best forecaster. After identifying the cause, the matching correction is: Evaluate ARIMA_PLUS or an appropriate time-series forecasting model before adopting an expensive generative approach. This matters here because during model-training validation on a Google Cloud experimental project, the following condition occurs: An….

Incorrect Answers

A:

Inspect task requirements for retrieval ranking, available query-document pairs, embedding similarity and labeled relevance examples. This examines embedding retrieval versus text classifier rather than the specific mechanism reported here. It is not the decisive first check for forecast trend and seasonality versus llm. The relevant evidence is: Inspect time stamps, autocorrelation, horizon, seasonal cycles and out-of-sample forecast metrics against traditional baselines.

B:

Determine whether failures arise from changing reference knowledge, stable behavior patterns or inadequate retrieval coverage. This examines foundation model prompt versus fine-tuning rather than the specific mechanism reported here. It is not the decisive first check for forecast trend and seasonality versus llm. The relevant evidence is: Inspect time stamps, autocorrelation, horizon, seasonal cycles and out-of-sample forecast metrics against traditional baselines. This matters here because during model-training validation on a Google Cloud experimental project, the following condition occurs: An….

D:

Compare pretraining benchmark scores, domain validation curves, trainable parameters and fine-tuning learning-rate choices. This examines fine-tuning catastrophic forgetting rather than the specific mechanism reported here. It is not the decisive first check for forecast trend and seasonality versus llm. The relevant evidence is: Inspect time stamps, autocorrelation, horizon, seasonal cycles and out-of-sample forecast metrics against traditional baselines. This matters here because during model-training validation on a Google Cloud experimental project, the following condition occurs: An….

Question 816

During a production ML inference release on Google Cloud, the following condition occurs: A real-time scoring service reads a managed feature view that has not synchronized since yesterday while the source BigQuery table contains recent transaction updates. The service must meet its agreed quality and latency requirements. What acceptance result verifies the intended behavior after remediation?

  1. Only a limited portion initially uses the candidate and a simulated regression triggers a tested rollback.
  2. The model loads at target concurrency without OOM and meets latency and cost targets.
  3. The morning surge meets availability and p95 latency objectives while off-peak cost remains documented.
  4. Authorized lookups return expected recent values and freshness alarms fire when synchronization falls behind.

Explanation

Correct Answer

D:

Authorized lookups return expected recent values and freshness alarms fire when synchronization falls behind. This is specific acceptance evidence for feature store freshness and online sync. Online feature serving requires an up-to-date synchronized view; source-table changes need not appear instantly. It confirms the controlled correction: Schedule and monitor approved Feature Store synchronization at the needed frequency, with stale-feature handling.

Incorrect Answers

A:

Only a limited portion initially uses the candidate and a simulated regression triggers a tested rollback. This result could demonstrate success for canary gradual traffic rollout but not for the stated feature store freshness and online sync issue. The direct test is: Authorized lookups return expected recent values and freshness alarms fire when synchronization falls behind.

B:

The model loads at target concurrency without OOM and meets latency and cost targets. This result could demonstrate success for gpu cost and memory fit for large embedding model but not for the stated feature store freshness and online sync issue. The direct test is: Authorized lookups return expected recent values and freshness alarms fire when synchronization falls behind.

C:

The morning surge meets availability and p95 latency objectives while off-peak cost remains documented. This result could demonstrate success for autoscaling minimum capacity versus cold load but not for the stated feature store freshness and online sync issue. The direct test is: Authorized lookups return expected recent values and freshness alarms fire when synchronization falls behind.

Question 817

At a Google Cloud analytics team preparing a governed pilot, the following condition occurs: A retailer has no trustworthy segment labels but attempts a supervised classifier with synthetic classes supplied by the sales team. The team must satisfy a reproducible low-code AI assessment. Which underlying engineering principle explains why the original approach fails?

  1. K-means discovers unlabeled similarity groups; fabricating labels would make supervised accuracy meaningless.
  2. A pretrained vision API may cover general image concepts, while specialized categories may need a custom model.
  3. In-database prediction avoids unnecessary bulk data transfer when model and features are already in BigQuery.
  4. Classification labels cannot directly represent unrestricted continuous outcomes; SQL regression models match numeric targets.

Explanation

Correct Answer

A:

K-means discovers unlabeled similarity groups; fabricating labels would make supervised accuracy meaningless. This explains the mechanism behind bigquery ml kmeans customer segmentation and why the described failure is possible. A discriminating investigation is: Check whether real class labels exist and examine the standardized input distributions and cluster stability under different seeds. The corresponding repair is: Train an appropriately scaled BigQuery ML KMEANS model and assess cluster cohesion, size and business interpretability.

Incorrect Answers

B:

A pretrained vision API may cover general image concepts, while specialized categories may need a custom model. This principle describes vision api pretrained image labels versus custom detection, which is not the cause of the observed bigquery ml kmeans customer segmentation failure. The controlling explanation is: K-means discovers unlabeled similarity groups; fabricating labels would make supervised accuracy meaningless.

C:

In-database prediction avoids unnecessary bulk data transfer when model and features are already in BigQuery. This principle describes bigquery ml sql batch prediction integration, which is not the cause of the observed bigquery ml kmeans customer segmentation failure. The controlling explanation is: K-means discovers unlabeled similarity groups; fabricating labels would make supervised accuracy meaningless.

D:

Classification labels cannot directly represent unrestricted continuous outcomes; SQL regression models match numeric targets. This principle describes bigquery ml linear regression for continuous value, which is not the cause of the observed bigquery ml kmeans customer segmentation failure. The controlling explanation is: K-means discovers unlabeled similarity groups; fabricating labels would make supervised accuracy meaningless.

Question 818

As the team tests a new governed AI release against adversarial and normal data, the following condition occurs: An online classifier uses a 24-hour account activity total while training created a seven-day total with the same column name. The fix must retain legitimate model usefulness and appropriate controls. Which TWO complementary choices jointly establish the evidence and appropriate correction? Select TWO.

  1. Present feature attributions as model-decision associations and use proper causal analysis before making intervention claims.
  2. Inspect cohort conversion metrics, assignment validity, minimum sample requirements and the release guardrail configuration.
  3. Add groundedness and factual correctness checks to a representative gen AI evaluation suite and tune retrieval or prompting accordingly.
  4. Compare live and offline feature SQL, window definitions, sample entity values and prediction-time cutoffs.
  5. Version a single feature contract and add automated parity tests to the preprocessing pipeline and serving workflow.

Explanation

Correct Answers

D:

Compare live and offline feature SQL, window definitions, sample entity values and prediction-time cutoffs. This inspects evidence relevant to training-serving feature skew detection. Identical feature names do not guarantee identical semantics; mismatched windows create training-serving skew. After identifying the cause, the matching correction is: Version a single feature contract and add automated parity tests to the preprocessing pipeline and serving workflow. This matters here because as the team tests a new governed AI release against adversarial and normal data, the following condition occurs: An….

E:

Version a single feature contract and add automated parity tests to the preprocessing pipeline and serving workflow. This is the targeted corrective response for training-serving feature skew detection. Identical feature names do not guarantee identical semantics; mismatched windows create training-serving skew. Verify the diagnosis using: Compare live and offline feature SQL, window definitions, sample entity values and prediction-time cutoffs. Confirm success using: For controlled entities the online and offline features match at equivalent cutoffs and quality recovers. This matters here because as the team tests a new governed AI release against adversarial and normal data, the following condition occurs: An….

Incorrect Answers

A:

Present feature attributions as model-decision associations and use proper causal analysis before making intervention claims. This is an action or diagnostic relevant to explainability versus causal claim, not one of the required two steps. In this scenario verify: Compare live and offline feature SQL, window definitions, sample entity values and prediction-time cutoffs. Then correct: Version a single feature contract and add automated parity tests to the preprocessing pipeline and serving workflow.

B:

Inspect cohort conversion metrics, assignment validity, minimum sample requirements and the release guardrail configuration. This is an action or diagnostic relevant to production incident rollback based on metrics, not one of the required two steps. In this scenario verify: Compare live and offline feature SQL, window definitions, sample entity values and prediction-time cutoffs. Then correct: Version a single feature contract and add automated parity tests to the preprocessing pipeline and serving workflow.

C:

Add groundedness and factual correctness checks to a representative gen AI evaluation suite and tune retrieval or prompting accordingly. This is an action or diagnostic relevant to generative ai factual grounding evaluation, not one of the required two steps. In this scenario verify: Compare live and offline feature SQL, window definitions, sample entity values and prediction-time cutoffs. Then correct: Version a single feature contract and add automated parity tests to the preprocessing pipeline and serving workflow.

Question 819

During a staged restoration test at a secondary serving deployment, the following condition occurs: A private inference service exists with proper invoker IAM, but a VPC client cannot connect because the private route and DNS mapping are missing. The team must prove an actual user-visible improvement without weakening access controls. Which TWO complementary choices jointly establish the evidence and appropriate correction? Select TWO.

  1. Inspect resolved endpoint address, authorized network routing, service access policy and connection logs.
  2. Choose a managed batch inference or scheduled container batch workflow instead of an always-on online endpoint.
  3. Inspect idle cost, model initialization time, min replicas, burst arrival, queueing and request deadlines.
  4. Test a supported post-training quantization or optimization path and release only if target quality remains acceptable.
  5. Configure the supported private DNS, network route and endpoint access path without granting public ingress.

Explanation

Correct Answers

A:

Inspect resolved endpoint address, authorized network routing, service access policy and connection logs. This inspects evidence relevant to private vpc routing failure for serving. Identity permission is not proof of network reachability; private serving also requires correct routing and name resolution. After identifying the cause, the matching correction is: Configure the supported private DNS, network route and endpoint access path without granting public ingress. This matters here because during a staged restoration test at a secondary serving deployment, the following condition occurs: A private….

E:

Configure the supported private DNS, network route and endpoint access path without granting public ingress. This is the targeted corrective response for private vpc routing failure for serving. Identity permission is not proof of network reachability; private serving also requires correct routing and name resolution. Verify the diagnosis using: Inspect resolved endpoint address, authorized network routing, service access policy and connection logs. Confirm success using: A test client on the approved network succeeds while an external probe remains blocked. This matters here because during a staged restoration test at a secondary serving deployment, the following condition occurs: A private….

Incorrect Answers

B:

Choose a managed batch inference or scheduled container batch workflow instead of an always-on online endpoint. This is an action or diagnostic relevant to batch versus online inference schedule, not one of the required two steps. In this scenario verify: Inspect resolved endpoint address, authorized network routing, service access policy and connection logs. Then correct: Configure the supported private DNS, network route and endpoint access path without granting public ingress.

C:

Inspect idle cost, model initialization time, min replicas, burst arrival, queueing and request deadlines. This is an action or diagnostic relevant to autoscaling minimum capacity versus cold load, not one of the required two steps. In this scenario verify: Inspect resolved endpoint address, authorized network routing, service access policy and connection logs. Then correct: Configure the supported private DNS, network route and endpoint access path without granting public ingress.

D:

Test a supported post-training quantization or optimization path and release only if target quality remains acceptable. This is an action or diagnostic relevant to inference optimization using quantization, not one of the required two steps. In this scenario verify: Inspect resolved endpoint address, authorized network routing, service access policy and connection logs. Then correct: Configure the supported private DNS, network route and endpoint access path without granting public ingress.

Question 820

As the team tests a new governed AI release against adversarial and normal data, the following condition occurs: A model deployment's business conversion rate falls during canary traffic, but the pipeline automatically continues rolling out because only container CPU is monitored. The fix must retain legitimate model usefulness and appropriate controls. Which evidence most directly establishes what is wrong with this AI or ML workflow?

  1. Compare time-stratified labeled performance, feature distribution stability and delayed ground truth coverage.
  2. Inspect live feature histograms, training baselines, data-capture completeness and drift metric thresholds.
  3. Inspect cohort conversion metrics, assignment validity, minimum sample requirements and the release guardrail configuration.
  4. Inspect input logs, sensitivity classifications, prompt assembly and approved data-processing boundaries.

Explanation

Correct Answer

C:

Inspect cohort conversion metrics, assignment validity, minimum sample requirements and the release guardrail configuration. This inspects evidence relevant to production incident rollback based on metrics. Resource health cannot substitute for model outcome quality; release decisions need explicit relevant monitoring signals. After identifying the cause, the matching correction is: Require validated business quality and technical reliability thresholds to continue rollout, with tested rollback on failure.

Incorrect Answers

A:

Compare time-stratified labeled performance, feature distribution stability and delayed ground truth coverage. This examines concept drift with stable feature statistics rather than the specific mechanism reported here. It is not the decisive first check for production incident rollback based on metrics. The relevant evidence is: Inspect cohort conversion metrics, assignment validity, minimum sample requirements and the release guardrail configuration.

B:

Inspect live feature histograms, training baselines, data-capture completeness and drift metric thresholds. This examines input data distribution drift alarm rather than the specific mechanism reported here. It is not the decisive first check for production incident rollback based on metrics. The relevant evidence is: Inspect cohort conversion metrics, assignment validity, minimum sample requirements and the release guardrail configuration.

D:

Inspect input logs, sensitivity classifications, prompt assembly and approved data-processing boundaries. This examines sensitive identifiers leaked to llm prompts rather than the specific mechanism reported here. It is not the decisive first check for production incident rollback based on metrics. The relevant evidence is: Inspect cohort conversion metrics, assignment validity, minimum sample requirements and the release guardrail configuration.

Professional Machine Learning Engineer Practice Tests