100% Latest & Updated Google Professional Machine Learning Engineer Practice Test Questions, Exam Dumps & Verified Answers!
30 Days Free Updates, Instant Download!
Question 110
A deployment has been completed and needs an acceptance test: Select an available Imagen image-generation model from Model Garden with controlled prompts. This intervention addresses: A marketing team needs concept illustrations described in text and has no base photos to classify or enhance. Which post-change check most directly proves the intended result?
- Review category-level precision and low-confidence manual queues.
- Inspect generated assets for prompt fidelity, unwanted text and brand compliance.
- Evaluate outputs against human-reviewed photo annotations and unsupported-claim rates.
- Check bilingual evaluation on a terminology-rich representative article set.
Explanation
Correct Answer
B:This is the appropriate post-change test: Inspect generated assets for prompt fidelity, unwanted text and brand compliance. The change under test is: Select an available Imagen image-generation model from Model Garden with controlled prompts. The original condition was: A marketing team needs concept illustrations described in text and has no base photos to classify or enhance. The expected behavior is: The team can draft original visuals that meet the specified concept.
Incorrect Answers
A:This check relates to another system: A retailer needs common object labels for new catalog photos without an internal visual classifier or labeled product corpus. It would not verify whether the current problem was fixed: A marketing team needs concept illustrations described in text and has no base photos to classify or enhance. The appropriate acceptance test is: Inspect generated assets for prompt fidelity, unwanted text and brand compliance.
C:This check relates to another system: An insurance claims tool needs a textual summary that combines a customer message with an attached claim photo. It would not verify whether the current problem was fixed: A marketing team needs concept illustrations described in text and has no base photos to classify or enhance. The appropriate acceptance test is: Inspect generated assets for prompt fidelity, unwanted text and brand compliance.
D:This check relates to another system: A travel website must translate brief help articles into French and Japanese and has no aligned text corpus for training a bespoke translation model. It would not verify whether the current problem was fixed: A marketing team needs concept illustrations described in text and has no base photos to classify or enhance. The appropriate acceptance test is: Inspect generated assets for prompt fidelity, unwanted text and brand compliance.
Question 111
After updating the service, engineers need convincing operational verification: Use an appropriate available Veo generative-video workflow with reviewed prompts. This intervention addresses: An advertising prototype needs brief new moving footage from a storyboard, not a caption describing an uploaded video. Which post-change check most directly proves the intended result?
- Benchmark representative requests on the deployment Region using fixed evaluation prompts.
- Check bilingual evaluation on a terminology-rich representative article set.
- Sample transcript word errors and sensitive-data handling by call type.
- Review frame continuity, safety and actual rendered output against the storyboard.
Explanation
Correct Answer
D:This is the appropriate post-change test: Review frame continuity, safety and actual rendered output against the storyboard. The change under test is: Use an appropriate available Veo generative-video workflow with reviewed prompts. The original condition was: An advertising prototype needs brief new moving footage from a storyboard, not a caption describing an uploaded video. The expected behavior is: The storyboard can become new synthesized video clips for editorial review.
Incorrect Answers
A:This check relates to another system: An engineering group needs a small low-latency text model but accidentally benchmarks only the largest expensive foundation model. It would not verify whether the current problem was fixed: An advertising prototype needs brief new moving footage from a storyboard, not a caption describing an uploaded video. The appropriate acceptance test is: Review frame continuity, safety and actual rendered output against the storyboard.
B:This check relates to another system: A travel website must translate brief help articles into French and Japanese and has no aligned text corpus for training a bespoke translation model. It would not verify whether the current problem was fixed: An advertising prototype needs brief new moving footage from a storyboard, not a caption describing an uploaded video. The appropriate acceptance test is: Review frame continuity, safety and actual rendered output against the storyboard.
C:This check relates to another system: A call center wants searchable transcripts of long recorded calls in supported languages without custom acoustic model training. It would not verify whether the current problem was fixed: An advertising prototype needs brief new moving footage from a storyboard, not a caption describing an uploaded video. The appropriate acceptance test is: Review frame continuity, safety and actual rendered output against the storyboard.
Question 112
The team wants an explicit pass/fail check after this implementation: Evaluate several compatible first-party and open or partner models in Model Garden. This intervention addresses: An engineering group needs a small low-latency text model but accidentally benchmarks only the largest expensive foundation model. Which post-change check most directly proves the intended result?
- Review frame continuity, safety and actual rendered output against the storyboard.
- Sample transcript word errors and sensitive-data handling by call type.
- Review category-level precision and low-confidence manual queues.
- Benchmark representative requests on the deployment Region using fixed evaluation prompts.
Explanation
Correct Answer
D:This is the appropriate post-change test: Benchmark representative requests on the deployment Region using fixed evaluation prompts. The change under test is: Evaluate several compatible first-party and open or partner models in Model Garden. The original condition was: An engineering group needs a small low-latency text model but accidentally benchmarks only the largest expensive foundation model. The expected behavior is: The team can select a model meeting quality thresholds at a lower serving cost.
Incorrect Answers
A:This check relates to another system: An advertising prototype needs brief new moving footage from a storyboard, not a caption describing an uploaded video. It would not verify whether the current problem was fixed: An engineering group needs a small low-latency text model but accidentally benchmarks only the largest expensive foundation model. The appropriate acceptance test is: Benchmark representative requests on the deployment Region using fixed evaluation prompts.
B:This check relates to another system: A call center wants searchable transcripts of long recorded calls in supported languages without custom acoustic model training. It would not verify whether the current problem was fixed: An engineering group needs a small low-latency text model but accidentally benchmarks only the largest expensive foundation model. The appropriate acceptance test is: Benchmark representative requests on the deployment Region using fixed evaluation prompts.
C:This check relates to another system: A retailer needs common object labels for new catalog photos without an internal visual classifier or labeled product corpus. It would not verify whether the current problem was fixed: An engineering group needs a small low-latency text model but accidentally benchmarks only the largest expensive foundation model. The appropriate acceptance test is: Benchmark representative requests on the deployment Region using fixed evaluation prompts.
Question 113
A deployment has been completed and needs an acceptance test: Build retrieval-grounded generation with access-controlled search and source citations. This intervention addresses: A company chatbot answers questions about frequently changing internal procedures but a pretrained LLM invents obsolete policy details. Which post-change check most directly proves the intended result?
- Sample transcript word errors and sensitive-data handling by call type.
- Check bilingual evaluation on a terminology-rich representative article set.
- Track unsafe response rate and false blocks over a representative red-team suite.
- Measure citation support and correctness on fresh policy questions with known answers.
Explanation
Correct Answer
D:This is the appropriate post-change test: Measure citation support and correctness on fresh policy questions with known answers. The change under test is: Build retrieval-grounded generation with access-controlled search and source citations. The original condition was: A company chatbot answers questions about frequently changing internal procedures but a pretrained LLM invents obsolete policy details. The expected behavior is: Answers can reference current retrieved procedures rather than relying on model memory.
Incorrect Answers
A:This check relates to another system: A call center wants searchable transcripts of long recorded calls in supported languages without custom acoustic model training. It would not verify whether the current problem was fixed: A company chatbot answers questions about frequently changing internal procedures but a pretrained LLM invents obsolete policy details. The appropriate acceptance test is: Measure citation support and correctness on fresh policy questions with known answers.
B:This check relates to another system: A travel website must translate brief help articles into French and Japanese and has no aligned text corpus for training a bespoke translation model. It would not verify whether the current problem was fixed: A company chatbot answers questions about frequently changing internal procedures but a pretrained LLM invents obsolete policy details. The appropriate acceptance test is: Measure citation support and correctness on fresh policy questions with known answers.
C:This check relates to another system: A child-focused learning bot generates some unsafe text despite a system instruction asking for safe answers. It would not verify whether the current problem was fixed: A company chatbot answers questions about frequently changing internal procedures but a pretrained LLM invents obsolete policy details. The appropriate acceptance test is: Measure citation support and correctness on fresh policy questions with known answers.
Question 114
After updating the service, engineers need convincing operational verification: Use supported caching for stable repeated prompt context with appropriate invalidation. This intervention addresses: A support assistant repeats a very large fixed policy prefix in every request, driving unnecessary processing time and tokens. Which post-change check most directly proves the intended result?
- Measure cache hit rate, p95 response time and billed token categories.
- Check bilingual evaluation on a terminology-rich representative article set.
- Review frame continuity, safety and actual rendered output against the storyboard.
- Evaluate outputs against human-reviewed photo annotations and unsupported-claim rates.
Explanation
Correct Answer
A:This is the appropriate post-change test: Measure cache hit rate, p95 response time and billed token categories. The change under test is: Use supported caching for stable repeated prompt context with appropriate invalidation. The original condition was: A support assistant repeats a very large fixed policy prefix in every request, driving unnecessary processing time and tokens. The expected behavior is: Repeated requests can reuse cached context where the provider supports it.
Incorrect Answers
B:This check relates to another system: A travel website must translate brief help articles into French and Japanese and has no aligned text corpus for training a bespoke translation model. It would not verify whether the current problem was fixed: A support assistant repeats a very large fixed policy prefix in every request, driving unnecessary processing time and tokens. The appropriate acceptance test is: Measure cache hit rate, p95 response time and billed token categories.
C:This check relates to another system: An advertising prototype needs brief new moving footage from a storyboard, not a caption describing an uploaded video. It would not verify whether the current problem was fixed: A support assistant repeats a very large fixed policy prefix in every request, driving unnecessary processing time and tokens. The appropriate acceptance test is: Measure cache hit rate, p95 response time and billed token categories.
D:This check relates to another system: An insurance claims tool needs a textual summary that combines a customer message with an attached claim photo. It would not verify whether the current problem was fixed: A support assistant repeats a very large fixed policy prefix in every request, driving unnecessary processing time and tokens. The appropriate acceptance test is: Measure cache hit rate, p95 response time and billed token categories.
Question 115
The team wants an explicit pass/fail check after this implementation: Apply supported safety settings and separate content-policy screening and human escalation. This intervention addresses: A child-focused learning bot generates some unsafe text despite a system instruction asking for safe answers. Which post-change check most directly proves the intended result?
- Measure citation support and correctness on fresh policy questions with known answers.
- Review category-level precision and low-confidence manual queues.
- Compare extracted invoice totals and supplier identifiers against adjudicated samples.
- Track unsafe response rate and false blocks over a representative red-team suite.
Explanation
Correct Answer
D:This is the appropriate post-change test: Track unsafe response rate and false blocks over a representative red-team suite. The change under test is: Apply supported safety settings and separate content-policy screening and human escalation. The original condition was: A child-focused learning bot generates some unsafe text despite a system instruction asking for safe answers. The expected behavior is: Unsafe requests can be blocked or routed for review more consistently.
Incorrect Answers
A:This check relates to another system: A company chatbot answers questions about frequently changing internal procedures but a pretrained LLM invents obsolete policy details. It would not verify whether the current problem was fixed: A child-focused learning bot generates some unsafe text despite a system instruction asking for safe answers. The appropriate acceptance test is: Track unsafe response rate and false blocks over a representative red-team suite.
B:This check relates to another system: A retailer needs common object labels for new catalog photos without an internal visual classifier or labeled product corpus. It would not verify whether the current problem was fixed: A child-focused learning bot generates some unsafe text despite a system instruction asking for safe answers. The appropriate acceptance test is: Track unsafe response rate and false blocks over a representative red-team suite.
C:This check relates to another system: An accounts-payable service receives variable-layout scanned invoices and needs supplier names, totals and line items without training OCR from scratch. It would not verify whether the current problem was fixed: A child-focused learning bot generates some unsafe text despite a system instruction asking for safe answers. The appropriate acceptance test is: Track unsafe response rate and false blocks over a representative red-team suite.
Question 116
An ML platform team is resolving 2 independent workstreams. Workstream A: A training feature query scans seven years of purchase logs although the model only needs the latest 90 days. Workstream B: A loan model has inflated test accuracy because it joins the customer's latest balance to examples from six months earlier. Select exactly TWO actions, one directly addressing each separate workstream.
- Run a Dataflow streaming pipeline using event-time windows and appropriate lateness policy.
- Validate types and mandatory columns during ingestion and quarantine incompatible files.
- Use a supported managed Spark pipeline with skew-aware partitioning and join strategy.
- Build event-time-correct historical feature joins from retained feature records.
- Apply correct date partition predicates and select only necessary fields.
Explanation
Correct Answers
D:Build event-time-correct historical feature joins from retained feature records. Check a sample of historical rows for features recorded after their label timestamp. Storing today's latest feature values is not enough to reconstruct the past.
E:Apply correct date partition predicates and select only necessary fields. Compare bytes processed and output-row equality before and after optimization. A predicate on an incompatible expression may not prune target partitions.
Incorrect Answers
A:Running a Dataflow streaming pipeline using event-time windows and appropriate lateness policy is relevant when a mobile-app stream delivers delayed purchase events and processing-time aggregation misassigns revenue to the wrong business hour; it does not establish the result required by this workstream.
B:The approach “Validate types and mandatory columns during ingestion and quarantine incompatible files” is relevant when an upstream app changes a numeric purchase amount column into a string and downstream feature training silently drops rows; it does not establish the result required by this workstream.
C:Using a supported managed Spark pipeline with skew-aware partitioning and join strategy is relevant when a 12 TB dataset requires joins with skewed keys and an in-memory pandas prototype runs out of memory; it does not establish the result required by this workstream.
Question 117
An ML engineering team must resolve this production problem: A mobile-app stream delivers delayed purchase events and processing-time aggregation misassigns revenue to the wrong business hour. Which action best addresses the actual technical requirement?
- Run a Dataflow streaming pipeline using event-time windows and appropriate lateness policy.
- Deidentify or tokenize protected fields before creating the shared training dataset.
- Build event-time-correct historical feature joins from retained feature records.
- Apply correct date partition predicates and select only necessary fields.
Explanation
Correct Answer
A:Run a Dataflow streaming pipeline using event-time windows and appropriate lateness policy. Compare replayed aggregates with authoritative hourly revenue totals. An unlimited lateness window increases retained state and processing costs.
Incorrect Answers
B:That measure applies when researchers need statistics from customer service logs but raw identifiers are restricted from broad analysis accounts. The required intervention here is different: Run a Dataflow streaming pipeline using event-time windows and appropriate lateness policy. The need is a mobile-app stream delivers delayed purchase events and processing-time aggregation misassigns revenue to the wrong business hour.
C:That measure applies when a loan model has inflated test accuracy because it joins the customer's latest balance to examples from six months earlier. The required intervention here is different: Run a Dataflow streaming pipeline using event-time windows and appropriate lateness policy. The need is a mobile-app stream delivers delayed purchase events and processing-time aggregation misassigns revenue to the wrong business hour.
D:That measure applies when a training feature query scans seven years of purchase logs although the model only needs the latest 90 days. The required intervention here is different: Run a Dataflow streaming pipeline using event-time windows and appropriate lateness policy. The need is a mobile-app stream delivers delayed purchase events and processing-time aggregation misassigns revenue to the wrong business hour.
Question 118
An ML platform team is resolving 2 independent workstreams. Workstream A: A pubsub-triggered enrichment job updates a customer feature twice whenever message delivery retries occur. Workstream B: An upstream app changes a numeric purchase amount column into a string and downstream feature training silently drops rows. Select exactly TWO actions, one directly addressing each separate workstream.
- Use rolling-origin or chronological evaluation windows that predict future dates from past data.
- Build event-time-correct historical feature joins from retained feature records.
- Implement idempotent event processing and stable deduplication keys in the pipeline.
- Validate types and mandatory columns during ingestion and quarantine incompatible files.
- Use a supported managed Spark pipeline with skew-aware partitioning and join strategy.
Explanation
Correct Answers
C:Implement idempotent event processing and stable deduplication keys in the pipeline. Replay duplicate events and compare final feature counts to the unique-event baseline. Delivery semantics alone do not guarantee exactly-once side effects in every external sink.
D:Validate types and mandatory columns during ingestion and quarantine incompatible files. Track rejected count, declared schema version and downstream row completeness. Coercing every malformed value may conceal source errors and alter statistics.
Incorrect Answers
A:Using rolling-origin or chronological evaluation windows that predict future dates from past data is relevant when a prediction service forecasts next week's demand but its accuracy was measured on randomly shuffled days; it does not establish the result required by this workstream.
B:Building event-time-correct historical feature joins from retained feature records is relevant when a loan model has inflated test accuracy because it joins the customer's latest balance to examples from six months earlier; it does not establish the result required by this workstream.
E:Using a supported managed Spark pipeline with skew-aware partitioning and join strategy is relevant when a 12 TB dataset requires joins with skewed keys and an in-memory pandas prototype runs out of memory; it does not establish the result required by this workstream.
Question 119
A Google Cloud change request describes the following issue: A 12 TB dataset requires joins with skewed keys and an in-memory pandas prototype runs out of memory. Which action best addresses the actual technical requirement?
- Build event-time-correct historical feature joins from retained feature records.
- Validate types and mandatory columns during ingestion and quarantine incompatible files.
- Use a supported managed Spark pipeline with skew-aware partitioning and join strategy.
- Implement idempotent event processing and stable deduplication keys in the pipeline.
Explanation
Correct Answer
C:Use a supported managed Spark pipeline with skew-aware partitioning and join strategy. Measure spill, straggler tasks and joined-row integrity on the distributed run. Hot keys and poor partition sizing can still stall large joins.
Incorrect Answers
A:That measure applies when a loan model has inflated test accuracy because it joins the customer's latest balance to examples from six months earlier. The required intervention here is different: Use a supported managed Spark pipeline with skew-aware partitioning and join strategy. The need is a 12 TB dataset requires joins with skewed keys and an in-memory pandas prototype runs out of memory.
B:That measure applies when an upstream app changes a numeric purchase amount column into a string and downstream feature training silently drops rows. The required intervention here is different: Use a supported managed Spark pipeline with skew-aware partitioning and join strategy. The need is a 12 TB dataset requires joins with skewed keys and an in-memory pandas prototype runs out of memory.
D:That measure applies when a pubsub-triggered enrichment job updates a customer feature twice whenever message delivery retries occur. The required intervention here is different: Use a supported managed Spark pipeline with skew-aware partitioning and join strategy. The need is a 12 TB dataset requires joins with skewed keys and an in-memory pandas prototype runs out of memory.
Question 120
An ML platform team is resolving 2 independent workstreams. Workstream A: A loan model has inflated test accuracy because it joins the customer's latest balance to examples from six months earlier. Workstream B: A prediction service forecasts next week's demand but its accuracy was measured on randomly shuffled days. Select exactly TWO actions, one directly addressing each separate workstream.
- Use a supported managed Spark pipeline with skew-aware partitioning and join strategy.
- Build event-time-correct historical feature joins from retained feature records.
- Validate types and mandatory columns during ingestion and quarantine incompatible files.
- Implement idempotent event processing and stable deduplication keys in the pipeline.
- Use rolling-origin or chronological evaluation windows that predict future dates from past data.
Explanation
Correct Answers
B:Build event-time-correct historical feature joins from retained feature records. Check a sample of historical rows for features recorded after their label timestamp. Storing today's latest feature values is not enough to reconstruct the past.
E:Use rolling-origin or chronological evaluation windows that predict future dates from past data. Backtest at several forecast origins and inspect uncertainty by season. A single short validation window may miss changing seasonal demand patterns.
Incorrect Answers
A:Using a supported managed Spark pipeline with skew-aware partitioning and join strategy is relevant when a 12 TB dataset requires joins with skewed keys and an in-memory pandas prototype runs out of memory; it does not establish the result required by this workstream.
C:The approach “Validate types and mandatory columns during ingestion and quarantine incompatible files” is relevant when an upstream app changes a numeric purchase amount column into a string and downstream feature training silently drops rows; it does not establish the result required by this workstream.
D:The approach “Implement idempotent event processing and stable deduplication keys in the pipeline” is relevant when a pubsub-triggered enrichment job updates a customer feature twice whenever message delivery retries occur; it does not establish the result required by this workstream.
Professional Machine Learning Engineer Practice Tests