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 18

Question 170

During a production ML inference release on Google Cloud, the following condition occurs: A public prediction API queues unbounded client retries during a traffic spike, creating a cascading latency and memory outage. The service must meet its agreed quality and latency requirements. What acceptance result verifies the intended behavior after remediation?

  1. The running endpoint resolves to a traceable evaluated package and cannot silently switch weights when an alias moves.
  2. Peak traffic is rate-limited predictably and the serving fleet recovers without an unbounded request backlog.
  3. Regional clients experience lower p95 end-to-end latency without changes to model compute quality.
  4. Latency and memory decline on target hardware while validation degradation stays within the approved threshold.

Explanation

Correct Answer

B:

Peak traffic is rate-limited predictably and the serving fleet recovers without an unbounded request backlog. This is specific acceptance evidence for predictive serving rate limit and backpressure. Unlimited retries amplify overload; backpressure and controlled concurrency are needed along with scaling. It confirms the controlled correction: Apply bounded concurrency, idempotent retries with exponential backoff and explicit overload responses.

Incorrect Answers

A:

The running endpoint resolves to a traceable evaluated package and cannot silently switch weights when an alias moves. This result could demonstrate success for model registry version provenance but not for the stated predictive serving rate limit and backpressure issue. The direct test is: Peak traffic is rate-limited predictably and the serving fleet recovers without an unbounded request backlog.

C:

Regional clients experience lower p95 end-to-end latency without changes to model compute quality. This result could demonstrate success for serving region proximity versus model compute but not for the stated predictive serving rate limit and backpressure issue. The direct test is: Peak traffic is rate-limited predictably and the serving fleet recovers without an unbounded request backlog.

D:

Latency and memory decline on target hardware while validation degradation stays within the approved threshold. This result could demonstrate success for inference optimization using quantization but not for the stated predictive serving rate limit and backpressure issue. The direct test is: Peak traffic is rate-limited predictably and the serving fleet recovers without an unbounded request backlog.

Question 171

During a production ML inference release on Google Cloud, the following condition occurs: A model server scales down to zero expensive workers overnight but cannot meet the first morning burst because loading weights takes several minutes. The service must meet its agreed quality and latency requirements. Which result is the strongest postchange evidence of success?

  1. Previously missed customer IDs retrieve the intended feature values without collapsing distinct leading-zero identifiers.
  2. The morning surge meets availability and p95 latency objectives while off-peak cost remains documented.
  3. The nightly job completes within the agreed SLA without idle real-time serving charges.
  4. The model loads at target concurrency without OOM and meets latency and cost targets.

Explanation

Correct Answer

B:

The morning surge meets availability and p95 latency objectives while off-peak cost remains documented. This is specific acceptance evidence for autoscaling minimum capacity versus cold load. Reactive scaling cannot instantly serve requests when cold model initialization is long. It confirms the controlled correction: Set a justified warm minimum or scheduled capacity and balance initialization overhead against idle cost.

Incorrect Answers

A:

Previously missed customer IDs retrieve the intended feature values without collapsing distinct leading-zero identifiers. This result could demonstrate success for feature online entity key mismatch but not for the stated autoscaling minimum capacity versus cold load issue. The direct test is: The morning surge meets availability and p95 latency objectives while off-peak cost remains documented.

C:

The nightly job completes within the agreed SLA without idle real-time serving charges. This result could demonstrate success for batch versus online inference schedule but not for the stated autoscaling minimum capacity versus cold load issue. The direct test is: The morning surge meets availability and p95 latency objectives while off-peak cost remains documented.

D:

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 autoscaling minimum capacity versus cold load issue. The direct test is: The morning surge meets availability and p95 latency objectives while off-peak cost remains documented.

Question 172

During a controlled replay of the data and model handoff to a second engineering team, the following condition occurs: A churn model joins historical feature rows to today's latest cancellation status, exposing a future outcome to prior prediction dates. The handoff must preserve privacy, lineage and testable behavior. Which TWO choices together provide the decisive diagnostic evidence and the most direct acceptance test? Select TWO.

  1. The notebook still trains on approved sample tables while restricted production datasets deny unauthorized reads.
  2. Compare entity IDs, key types, online feature-view definitions, offline joins and records for the same verified customer.
  3. Inspect feature event timestamps, prediction cutoffs, feature joins and any post-cancellation administrative flags.
  4. Each candidate's reported metrics resolve to its actual artifact and replayable configuration without manual lookups.
  5. For every row the feature version predates its prediction cutoff and prospective holdout quality is realistic.

Explanation

Correct Answers

C:

Inspect feature event timestamps, prediction cutoffs, feature joins and any post-cancellation administrative flags. This inspects evidence relevant to point-in-time feature leakage prevention. The globally latest value may contain future information that could not have informed the original prediction. After identifying the cause, the matching correction is: Rebuild offline training data with point-in-time eligible joins and remove features available only after each prediction moment.

E:

For every row the feature version predates its prediction cutoff and prospective holdout quality is realistic. This is specific acceptance evidence for point-in-time feature leakage prevention. The globally latest value may contain future information that could not have informed the original prediction. It confirms the controlled correction: Rebuild offline training data with point-in-time eligible joins and remove features available only after each prediction moment.

Incorrect Answers

A:

The notebook still trains on approved sample tables while restricted production datasets deny unauthorized reads. This relates to colab enterprise notebook data access scope but does not establish the requested before-and-after evidence. The actual diagnostic is: Inspect feature event timestamps, prediction cutoffs, feature joins and any post-cancellation administrative flags. The specific acceptance result is: For every row the feature version predates its prediction cutoff and prospective holdout quality is realistic.

B:

Compare entity IDs, key types, online feature-view definitions, offline joins and records for the same verified customer. This relates to feature store online and offline entity alignment but does not establish the requested before-and-after evidence. The actual diagnostic is: Inspect feature event timestamps, prediction cutoffs, feature joins and any post-cancellation administrative flags. The specific acceptance result is: For every row the feature version predates its prediction cutoff and prospective holdout quality is realistic.

D:

Each candidate's reported metrics resolve to its actual artifact and replayable configuration without manual lookups. This relates to pipeline experiment versus notebook-only tracking but does not establish the requested before-and-after evidence. The actual diagnostic is: Inspect feature event timestamps, prediction cutoffs, feature joins and any post-cancellation administrative flags. The specific acceptance result is: For every row the feature version predates its prediction cutoff and prospective holdout quality is realistic.

Question 173

During a controlled replay of the data and model handoff to a second engineering team, the following condition occurs: An image labeling task sends full customer identity and contact records to external annotators although only product photographs need annotation. The handoff must preserve privacy, lineage and testable behavior. Which configuration or procedure is the most appropriate correction?

  1. Quarantine invalid spikes and fit a robust transformation using training data only, then apply it consistently at inference.
  2. Choose a compatible Model Garden model/runtime or perform a validated checkpoint conversion before training.
  3. Rewrite the filter to permit effective partition pruning while preserving its requested date boundaries.
  4. Redact or tokenize unnecessary identifying fields before labeling and limit access to approved mappings.

Explanation

Correct Answer

D:

Redact or tokenize unnecessary identifying fields before labeling and limit access to approved mappings. This is the targeted corrective response for data minimization for labeling inputs. Protecting data in transit does not excuse sending unnecessary sensitive attributes to a third-party workflow. Verify the diagnosis using: Inspect the labeling payload schema, data classification, annotator roles and fields needed to complete the labeling task. Confirm success using: Annotators can complete validated labels without seeing identifying data and the project retains compliant traceability. This matters here because during a controlled replay of the data and model handoff to a second engineering team, the following condition….

Incorrect Answers

A:

Quarantine invalid spikes and fit a robust transformation using training data only, then apply it consistently at inference. This action addresses outlier handling before feature standardization, not the failure in the described data minimization for labeling inputs case. The targeted correction is: Redact or tokenize unnecessary identifying fields before labeling and limit access to approved mappings.

B:

Choose a compatible Model Garden model/runtime or perform a validated checkpoint conversion before training. This action addresses model garden jax accelerator compatibility, not the failure in the described data minimization for labeling inputs case. The targeted correction is: Redact or tokenize unnecessary identifying fields before labeling and limit access to approved mappings.

C:

Rewrite the filter to permit effective partition pruning while preserving its requested date boundaries. This action addresses bigquery partition filter cost optimization, not the failure in the described data minimization for labeling inputs case. The targeted correction is: Redact or tokenize unnecessary identifying fields before labeling and limit access to approved mappings.

Question 174

During a staged restoration test at a secondary serving deployment, the following condition occurs: An online transformer processes one request per GPU kernel launch and throughput stays low despite steady high-frequency traffic. The team must prove an actual user-visible improvement without weakening access controls. Which TWO selections represent the correct investigation and remediation together? Select TWO.

  1. Implement the required serving contract and make the container report readiness only after loading model assets.
  2. Preserve stable canonical identifier types end-to-end and backfill only the affected feature records.
  3. Inspect batch size, kernel utilization, per-request overhead, queue wait and latency percentiles.
  4. Enable compatible dynamic batching with a bounded wait time and benchmark throughput against p95 latency.
  5. Compare request rate, queue depth, GPU utilization, per-instance concurrency and p95 response latency.

Explanation

Correct Answers

C:

Inspect batch size, kernel utilization, per-request overhead, queue wait and latency percentiles. This inspects evidence relevant to dynamic batch serving for gpu saturation. Bundling compatible requests can amortize GPU launch overhead, but excessive batch waiting raises response times. After identifying the cause, the matching correction is: Enable compatible dynamic batching with a bounded wait time and benchmark throughput against p95 latency.

D:

Enable compatible dynamic batching with a bounded wait time and benchmark throughput against p95 latency. This is the targeted corrective response for dynamic batch serving for gpu saturation. Bundling compatible requests can amortize GPU launch overhead, but excessive batch waiting raises response times. Verify the diagnosis using: Inspect batch size, kernel utilization, per-request overhead, queue wait and latency percentiles. Confirm success using: Requests per second increase while tail latency stays inside the service objective.

Incorrect Answers

A:

Implement the required serving contract and make the container report readiness only after loading model assets. This is an action or diagnostic relevant to inference container health interface mismatch, not one of the required two steps. In this scenario verify: Inspect batch size, kernel utilization, per-request overhead, queue wait and latency percentiles. Then correct: Enable compatible dynamic batching with a bounded wait time and benchmark throughput against p95 latency.

B:

Preserve stable canonical identifier types end-to-end and backfill only the affected feature records. This is an action or diagnostic relevant to feature online entity key mismatch, not one of the required two steps. In this scenario verify: Inspect batch size, kernel utilization, per-request overhead, queue wait and latency percentiles. Then correct: Enable compatible dynamic batching with a bounded wait time and benchmark throughput against p95 latency.

E:

Compare request rate, queue depth, GPU utilization, per-instance concurrency and p95 response latency. This is an action or diagnostic relevant to throughput-based autoscaling versus cpu metric, not one of the required two steps. In this scenario verify: Inspect batch size, kernel utilization, per-request overhead, queue wait and latency percentiles. Then correct: Enable compatible dynamic batching with a bounded wait time and benchmark throughput against p95 latency.

Question 175

During a production ML inference release on Google Cloud, the following condition occurs: A new image classifier has only offline test results, yet engineers plan to replace every customer prediction immediately after deployment. The service must meet its agreed quality and latency requirements. Which configuration or procedure is the most appropriate correction?

  1. Set a justified warm minimum or scheduled capacity and balance initialization overhead against idle cost.
  2. Roll out a small canary percentage first, monitor error and quality guardrails and expand only after acceptance.
  3. Benchmark a Cloud Run custom container against a managed Agent Platform endpoint and choose for workload fit.
  4. Promote an explicitly approved model version from Agent Platform Model Registry and record deployment lineage.

Explanation

Correct Answer

B:

Roll out a small canary percentage first, monitor error and quality guardrails and expand only after acceptance. This is the targeted corrective response for canary gradual traffic rollout. Progressive traffic exposure reduces blast radius compared with instant replacement of all production serving. Verify the diagnosis using: Inspect traffic-splitting capabilities, release guardrails, rollback triggers and representative online metrics. Confirm success using: Only a limited portion initially uses the candidate and a simulated regression triggers a tested rollback.

Incorrect Answers

A:

Set a justified warm minimum or scheduled capacity and balance initialization overhead against idle cost. This action addresses autoscaling minimum capacity versus cold load, not the failure in the described canary gradual traffic rollout case. The targeted correction is: Roll out a small canary percentage first, monitor error and quality guardrails and expand only after acceptance.

C:

Benchmark a Cloud Run custom container against a managed Agent Platform endpoint and choose for workload fit. This action addresses cloud run versus managed inference endpoint, not the failure in the described canary gradual traffic rollout case. The targeted correction is: Roll out a small canary percentage first, monitor error and quality guardrails and expand only after acceptance.

D:

Promote an explicitly approved model version from Agent Platform Model Registry and record deployment lineage. This action addresses model registry version provenance, not the failure in the described canary gradual traffic rollout case. The targeted correction is: Roll out a small canary percentage first, monitor error and quality guardrails and expand only after acceptance.

Question 176

During a regulated MLOps pipeline deployment, the following condition occurs: An automated retraining job overwrites the previous serving model before verifying quality under the most recent data distribution. The workflow must be repeatable and protect the last approved production model. Which targeted engineering change most directly addresses the failure?

  1. Create immutable versioned input snapshots or partition manifests and record them in training metadata.
  2. Use supported Ray retry policies and durable checkpoints at suitable stage boundaries instead of relying only on worker memory.
  3. Train only on adequately matured labeled cohorts and monitor recent feature drift separately until outcomes arrive.
  4. Retain the last healthy model and require canary validation before switching all prediction traffic.

Explanation

Correct Answer

D:

Retain the last healthy model and require canary validation before switching all prediction traffic. This is the targeted corrective response for continuous training rollback protection. Continuous training should not imply uncontrolled continuous deployment of every new model. Verify the diagnosis using: Review champion model retention, candidate metrics, change approvals and automatic rollback triggers. Confirm success using: An intentionally failing candidate is rolled back promptly and the previous version remains serving.

Incorrect Answers

A:

Create immutable versioned input snapshots or partition manifests and record them in training metadata. This action addresses data version reproducibility in ct jobs, not the failure in the described continuous training rollback protection case. The targeted correction is: Retain the last healthy model and require canary validation before switching all prediction traffic.

B:

Use supported Ray retry policies and durable checkpoints at suitable stage boundaries instead of relying only on worker memory. This action addresses ray distributed task failure isolation, not the failure in the described continuous training rollback protection case. The targeted correction is: Retain the last healthy model and require canary validation before switching all prediction traffic.

C:

Train only on adequately matured labeled cohorts and monitor recent feature drift separately until outcomes arrive. This action addresses scheduled retraining on delayed ground truth, not the failure in the described continuous training rollback protection case. The targeted correction is: Retain the last healthy model and require canary validation before switching all prediction traffic.

Question 177

While preparing a production training workflow on Gemini Enterprise Agent Platform, the following condition occurs: A retailer asks for next-week order value as a dollar amount, but the prototype returns yes/no purchase classes without estimated magnitude. Engineers need an auditable, resource-aware correction to the issue. Which diagnostic check should be performed FIRST to isolate the described cause?

  1. Inspect shutdown events, checkpoint directory, durable artifact transfers and restoration logic after simulated interruptions.
  2. Inspect required output type, business loss, sample labels, and the configured model objective and metric.
  3. Determine whether failures arise from changing reference knowledge, stable behavior patterns or inadequate retrieval coverage.
  4. Inspect task requirements for retrieval ranking, available query-document pairs, embedding similarity and labeled relevance examples.

Explanation

Correct Answer

B:

Inspect required output type, business loss, sample labels, and the configured model objective and metric. This inspects evidence relevant to regression versus classification product choice. Categorical classification is not an estimator of an unrestricted continuous order value. After identifying the cause, the matching correction is: Select a regression architecture and matching numeric loss rather than using a binary classifier for amounts.

Incorrect Answers

A:

Inspect shutdown events, checkpoint directory, durable artifact transfers and restoration logic after simulated interruptions. This examines training checkpoint persistence across interruptions rather than the specific mechanism reported here. It is not the decisive first check for regression versus classification product choice. The relevant evidence is: Inspect required output type, business loss, sample labels, and the configured model objective and metric.

C:

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 regression versus classification product choice. The relevant evidence is: Inspect required output type, business loss, sample labels, and the configured model objective and metric.

D:

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 regression versus classification product choice. The relevant evidence is: Inspect required output type, business loss, sample labels, and the configured model objective and metric.

Question 178

Within a jointly maintained Google Cloud data science workspace, the following condition occurs: A team uses manual spreadsheet rows for hyperparameter results and loses connections between training metrics and produced checkpoint artifacts. Collaborators must avoid expanding data access or breaking experiment reproducibility. Which evidence most directly establishes what is wrong with this AI or ML workflow?

  1. Inspect random seeds, sample ordering, framework determinism settings and environment versions recorded per run.
  2. Inspect judge rubric prompts, evaluation dataset balance, repeated ratings, human benchmark and uncertainty in pairwise preferences.
  3. Inspect pipeline execution IDs, experiment tracking records, artifact URIs and the link between metrics and saved models.
  4. Examine source locations, batch cadence, SQL transform complexity, streaming latency requirements and operational burden.

Explanation

Correct Answer

C:

Inspect pipeline execution IDs, experiment tracking records, artifact URIs and the link between metrics and saved models. This inspects evidence relevant to pipeline experiment versus notebook-only tracking. Trial metrics and artifact provenance must be linked to support valid experiment selection and release audits. After identifying the cause, the matching correction is: Use Agent Platform Experiments and pipeline metadata to track each trial's parameters, measurements and output artifacts.

Incorrect Answers

A:

Inspect random seeds, sample ordering, framework determinism settings and environment versions recorded per run. This examines experiment random seed and data order variance rather than the specific mechanism reported here. It is not the decisive first check for pipeline experiment versus notebook-only tracking. The relevant evidence is: Inspect pipeline execution IDs, experiment tracking records, artifact URIs and the link between metrics and saved models.

B:

Inspect judge rubric prompts, evaluation dataset balance, repeated ratings, human benchmark and uncertainty in pairwise preferences. This examines llm-as-a-judge rubric reliability rather than the specific mechanism reported here. It is not the decisive first check for pipeline experiment versus notebook-only tracking. The relevant evidence is: Inspect pipeline execution IDs, experiment tracking records, artifact URIs and the link between metrics and saved models.

D:

Examine source locations, batch cadence, SQL transform complexity, streaming latency requirements and operational burden. This examines bigquery versus dataflow preparation selection rather than the specific mechanism reported here. It is not the decisive first check for pipeline experiment versus notebook-only tracking. The relevant evidence is: Inspect pipeline execution IDs, experiment tracking records, artifact URIs and the link between metrics and saved models.

Question 179

During a regulated MLOps pipeline deployment, the following condition occurs: A fraud classifier is retrained every night despite true chargeback outcomes arriving weeks later, producing highly incomplete recent labels. The workflow must be repeatable and protect the last approved production model. Which evidence most directly establishes what is wrong with this AI or ML workflow?

  1. Inspect Cloud Build substitutions, commit SHAs, model registry versions, signed image digests and approval records.
  2. Inspect label arrival times, cohort maturity, feedback collection coverage and effective training cutoffs.
  3. Inspect Airflow retries, write mode, batch IDs, target partitions and commit logs.
  4. Inspect request schema, container route behavior, test fixtures and CI stage dependency graph.

Explanation

Correct Answer

B:

Inspect label arrival times, cohort maturity, feedback collection coverage and effective training cutoffs. This inspects evidence relevant to scheduled retraining on delayed ground truth. Unobserved outcomes cannot serve as reliable supervised labels regardless of the desired retraining schedule. After identifying the cause, the matching correction is: Train only on adequately matured labeled cohorts and monitor recent feature drift separately until outcomes arrive.

Incorrect Answers

A:

Inspect Cloud Build substitutions, commit SHAs, model registry versions, signed image digests and approval records. This examines cloud build release version promotion rather than the specific mechanism reported here. It is not the decisive first check for scheduled retraining on delayed ground truth. The relevant evidence is: Inspect label arrival times, cohort maturity, feedback collection coverage and effective training cutoffs.

C:

Inspect Airflow retries, write mode, batch IDs, target partitions and commit logs. This examines airflow task idempotency on retry rather than the specific mechanism reported here. It is not the decisive first check for scheduled retraining on delayed ground truth. The relevant evidence is: Inspect label arrival times, cohort maturity, feedback collection coverage and effective training cutoffs.

D:

Inspect request schema, container route behavior, test fixtures and CI stage dependency graph. This examines continuous delivery integration contract tests rather than the specific mechanism reported here. It is not the decisive first check for scheduled retraining on delayed ground truth. The relevant evidence is: Inspect label arrival times, cohort maturity, feedback collection coverage and effective training cutoffs.

Question 180

During a regulated MLOps pipeline deployment, the following condition occurs: A scheduled training pipeline fails during feature validation but the team discovers the failure only when customers complain about stale predictions. The workflow must be repeatable and protect the last approved production model. Which technical explanation best accounts for the reported behavior?

  1. Continuous training should not imply uncontrolled continuous deployment of every new model.
  2. Orchestration without failure notification can leave models stale even when the last successful deployment was correct.
  3. Model weights alone do not capture the data lineage necessary to reproduce and investigate predictions.
  4. Continuous training needs statistically and operationally justified policy thresholds rather than reactive promotion of noise.

Explanation

Correct Answer

B:

Orchestration without failure notification can leave models stale even when the last successful deployment was correct. This explains the mechanism behind pipeline notification alert on silent failure and why the described failure is possible. A discriminating investigation is: Inspect task execution states, failed-job logs, alert routing, pipeline-age metrics and owner escalation settings. The corresponding repair is: Configure failure and stale-model alerts tied to actionable run-status and freshness service objectives.

Incorrect Answers

A:

Continuous training should not imply uncontrolled continuous deployment of every new model. This principle describes continuous training rollback protection, which is not the cause of the observed pipeline notification alert on silent failure failure. The controlling explanation is: Orchestration without failure notification can leave models stale even when the last successful deployment was correct.

C:

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

D:

Continuous training needs statistically and operationally justified policy thresholds rather than reactive promotion of noise. This principle describes canary failure feedback into ct policy, which is not the cause of the observed pipeline notification alert on silent failure failure. The controlling explanation is: Orchestration without failure notification can leave models stale even when the last successful deployment was correct.

Professional Machine Learning Engineer Practice Tests