Exam Bundle

A training role cannot approve its own model package and a designated release role can promote a reviewed version. This would verify a different operation: Separate training, model approval and production deployment identities with explicit role and approval boundaries. That separate action addresses: A machine learning pipeline uses a single IAM role to train models, approve registry packages and update production endpoints. It cannot establish the required Westhaven Retail correction for: Production predictions fail silently during a partial downstream feature outage while endpoint uptime remains normal. The decisive acceptance check is instead: A simulated stale feature feed triggers a dependency alarm even though the SageMaker endpoint remains available.

A tagged cost breakdown identifies the new resource responsible for the measured increase and supports an accountable remediation. This would verify a different operation: Use AWS Cost Explorer and properly applied resource tags to compare endpoint, training and storage cost changes over time. That separate action addresses: An operations team tracks aggregate monthly AWS charges but cannot isolate a sudden cost increase to new model endpoints. It cannot establish the required Westhaven Retail correction for: Production predictions fail silently during a partial downstream feature outage while endpoint uptime remains normal. The decisive acceptance check is instead: A simulated stale feature feed triggers a dependency alarm even though the SageMaker endpoint remains available.

The nightly workload still meets its completion target while cost reports show reduced idle serving hours. This would verify a different operation: Measure invocation patterns and replace unnecessary continuous capacity with supported scheduled batch or appropriately scaled alternatives. That separate action addresses: An inference workload maintains two always-on high-memory SageMaker endpoints for models invoked only during a brief nightly window. It cannot establish the required Westhaven Retail correction for: Production predictions fail silently during a partial downstream feature outage while endpoint uptime remains normal. The decisive acceptance check is instead: A simulated stale feature feed triggers a dependency alarm even though the SageMaker endpoint remains available.

Pin a governed immutable model artifact version and validate checksums or signatures through the approved release process. It does not directly address the Granite SaaS case: Production predictions fail silently during a partial downstream feature outage while endpoint uptime remains normal. The relevant acceptance criterion for this case is: A simulated stale feature feed triggers a dependency alarm even though the SageMaker endpoint remains available.

Limit the SageMaker execution role and bucket policy to the approved data prefix and required object operations. It does not directly address the Granite SaaS case: Production predictions fail silently during a partial downstream feature outage while endpoint uptime remains normal. The relevant acceptance criterion for this case is: A simulated stale feature feed triggers a dependency alarm even though the SageMaker endpoint remains available.

Use approved private subnets, security groups and VPC endpoints where supported to restrict unnecessary network egress. It does not directly address the Granite SaaS case: Production predictions fail silently during a partial downstream feature outage while endpoint uptime remains normal. The relevant acceptance criterion for this case is: A simulated stale feature feed triggers a dependency alarm even though the SageMaker endpoint remains available.

A representative slow request shows the delay attributed to one documented processing stage rather than undifferentiated total time. This would verify a different operation: Instrument authorized distributed traces or structured per-stage timings, using AWS X-Ray where supported, to isolate bottlenecks. That separate action addresses: An ML API has intermittent latency spikes but CloudWatch cannot reveal whether the delay occurs in preprocessing or downstream inference. It cannot establish the required Granite SaaS correction for: Production predictions fail silently during a partial downstream feature outage while endpoint uptime remains normal. The decisive acceptance check is instead: A simulated stale feature feed triggers a dependency alarm even though the SageMaker endpoint remains available.

New output artifacts have the documented KMS key metadata and an attempted unencrypted write is rejected by bucket policy. This would verify a different operation: Require appropriate SSE-KMS settings on the output bucket and ensure the training role is authorized to use the chosen key. That separate action addresses: A model training job writes regulated outputs to an unencrypted S3 location despite a documented data-at-rest encryption standard. It cannot establish the required Granite SaaS correction for: Production predictions fail silently during a partial downstream feature outage while endpoint uptime remains normal. The decisive acceptance check is instead: A simulated stale feature feed triggers a dependency alarm even though the SageMaker endpoint remains available.

Objects past the retention window transition or expire as intended while protected evidence remains retrievable. This would verify a different operation: Apply reviewed S3 lifecycle transition and expiration policies aligned to audit retention and replay requirements. That separate action addresses: S3 inference request captures accumulate indefinitely in a costly storage class even after their approved monitoring retention period expires. It cannot establish the required Granite SaaS correction for: Production predictions fail silently during a partial downstream feature outage while endpoint uptime remains normal. The decisive acceptance check is instead: A simulated stale feature feed triggers a dependency alarm even though the SageMaker endpoint remains available.

Enforce protected build contexts, isolated untrusted code execution and restricted deployment credentials available only after review. It does not directly address the Bluehaven Transit case: A post-deployment bias analysis aggregates only predictions while never comparing their distribution with relevant protected-group outcomes. The relevant acceptance criterion for this case is: A controlled disparity case is surfaced by the intended bias metric and the report records sample sizes and review limitations.

Review artifact location, cache or image-layer design and choose a supported reuse strategy consistent with version integrity. It does not directly address the Bluehaven Transit case: A post-deployment bias analysis aggregates only predictions while never comparing their distribution with relevant protected-group outcomes. The relevant acceptance criterion for this case is: A controlled disparity case is surfaced by the intended bias metric and the report records sample sizes and review limitations.

Use pipeline-dependent training resource lifecycle controls with completion-triggered teardown and failure handling. It does not directly address the Bluehaven Transit case: A post-deployment bias analysis aggregates only predictions while never comparing their distribution with relevant protected-group outcomes. The relevant acceptance criterion for this case is: A controlled disparity case is surfaced by the intended bias metric and the report records sample sizes and review limitations.

A simulated untrusted pull request cannot assume the production release role and an approved release still deploys successfully. This would verify a different operation: Enforce protected build contexts, isolated untrusted code execution and restricted deployment credentials available only after review. That separate action addresses: An automated ML CI/CD runner accepts source contributions from untrusted forks while holding direct privileges to production SageMaker resources. It cannot establish the required Bluehaven Transit correction for: A post-deployment bias analysis aggregates only predictions while never comparing their distribution with relevant protected-group outcomes. The decisive acceptance check is instead: A controlled disparity case is surfaced by the intended bias metric and the report records sample sizes and review limitations.

A code scan finds no static secret and the notebook's execution role can use only its intended dataset resources. This would verify a different operation: Replace embedded access keys with authorized IAM roles and secure short-lived credential delivery for the notebook environment. That separate action addresses: A SageMaker notebook user stores long-lived AWS credentials in the notebook source so other project members can inspect them. It cannot establish the required Bluehaven Transit correction for: A post-deployment bias analysis aggregates only predictions while never comparing their distribution with relevant protected-group outcomes. The decisive acceptance check is instead: A controlled disparity case is surfaced by the intended bias metric and the report records sample sizes and review limitations.

A training role cannot approve its own model package and a designated release role can promote a reviewed version. This would verify a different operation: Separate training, model approval and production deployment identities with explicit role and approval boundaries. That separate action addresses: A machine learning pipeline uses a single IAM role to train models, approve registry packages and update production endpoints. It cannot establish the required Bluehaven Transit correction for: A post-deployment bias analysis aggregates only predictions while never comparing their distribution with relevant protected-group outcomes. The decisive acceptance check is instead: A controlled disparity case is surfaced by the intended bias metric and the report records sample sizes and review limitations.

Choose a suitable minimum serving capacity or compatible hosting method and validate the idle-cost versus cold-start tradeoff. It does not directly address the Brookside Manufacturing case: A post-deployment bias analysis aggregates only predictions while never comparing their distribution with relevant protected-group outcomes. The relevant acceptance criterion for this case is: A controlled disparity case is surfaced by the intended bias metric and the report records sample sizes and review limitations.

Use AWS Cost Explorer and properly applied resource tags to compare endpoint, training and storage cost changes over time. It does not directly address the Brookside Manufacturing case: A post-deployment bias analysis aggregates only predictions while never comparing their distribution with relevant protected-group outcomes. The relevant acceptance criterion for this case is: A controlled disparity case is surfaced by the intended bias metric and the report records sample sizes and review limitations.

Inspect SageMaker service quotas and scaling target limits and request justified quota changes before increasing traffic. It does not directly address the Brookside Manufacturing case: A post-deployment bias analysis aggregates only predictions while never comparing their distribution with relevant protected-group outcomes. The relevant acceptance criterion for this case is: A controlled disparity case is surfaced by the intended bias metric and the report records sample sizes and review limitations.

A cost model using observed utilization demonstrates the break-even point and documents the selected financial tradeoff. This would verify a different operation: Compare actual steady-state resource usage and purchasing alternatives before choosing a suitable commitment or On-Demand strategy. That separate action addresses: A team purchases reserved inference capacity without measuring whether its utilization is predictable enough to offset commitment cost. It cannot establish the required Brookside Manufacturing correction for: A post-deployment bias analysis aggregates only predictions while never comparing their distribution with relevant protected-group outcomes. The decisive acceptance check is instead: A controlled disparity case is surfaced by the intended bias metric and the report records sample sizes and review limitations.

A controlled endpoint configuration change yields an attributable CloudTrail event retained at the authorized destination. This would verify a different operation: Enable appropriate organization or account CloudTrail trails with secure retention and access controls for relevant management events. That separate action addresses: A security auditor cannot trace who modified a SageMaker endpoint because AWS CloudTrail event retention and organization trails were never configured. It cannot establish the required Brookside Manufacturing correction for: A post-deployment bias analysis aggregates only predictions while never comparing their distribution with relevant protected-group outcomes. The decisive acceptance check is instead: A controlled disparity case is surfaced by the intended bias metric and the report records sample sizes and review limitations.

A representative interruptible run resumes safely and its measured total cost is below the approved On-Demand baseline. This would verify a different operation: Evaluate SageMaker Managed Spot Training with checkpointing and compare interruption risk and effective completion cost. That separate action addresses: A training fleet uses expensive On-Demand instances for fault-tolerant short experiments without considering managed Spot capacity. It cannot establish the required Brookside Manufacturing correction for: A post-deployment bias analysis aggregates only predictions while never comparing their distribution with relevant protected-group outcomes. The decisive acceptance check is instead: A controlled disparity case is surfaced by the intended bias metric and the report records sample sizes and review limitations.

Add model-specific output-validity checks and application error metrics alongside endpoint health measurements. It does not directly address the Harborline Logistics case: An inference workload maintains two always-on high-memory SageMaker endpoints for models invoked only during a brief nightly window. The relevant acceptance criterion for this case is: The nightly workload still meets its completion target while cost reports show reduced idle serving hours.

Publish an approved data subset or governed view and grant read access only to the required machine learning project roles. It does not directly address the Harborline Logistics case: An inference workload maintains two always-on high-memory SageMaker endpoints for models invoked only during a brief nightly window. The relevant acceptance criterion for this case is: The nightly workload still meets its completion target while cost reports show reduced idle serving hours.

Create monitored business slices and evaluate approved per-segment metrics and bias indicators using retained ground-truth labels. It does not directly address the Harborline Logistics case: An inference workload maintains two always-on high-memory SageMaker endpoints for models invoked only during a brief nightly window. The relevant acceptance criterion for this case is: The nightly workload still meets its completion target while cost reports show reduced idle serving hours.

The monitoring report labels distribution shift and actual error metrics distinctly and does not infer accuracy loss from drift alone. This would verify a different operation: Distinguish covariate drift from verified performance degradation and evaluate outcome labels and model-quality metrics separately. That separate action addresses: An operations dashboard treats a rising model input drift score as proof that prediction correctness already decreased without label evidence. It cannot establish the required Harborline Logistics correction for: An inference workload maintains two always-on high-memory SageMaker endpoints for models invoked only during a brief nightly window. The decisive acceptance check is instead: The nightly workload still meets its completion target while cost reports show reduced idle serving hours.

A deliberately vulnerable test image is blocked and a compliant approved digest reaches the deployment stage. This would verify a different operation: Integrate ECR image scanning and an enforced reviewed vulnerability acceptance gate before deploying the immutable image digest. That separate action addresses: A training pipeline publishes an unscanned container image with known vulnerable dependencies directly to the production ECR repository. It cannot establish the required Harborline Logistics correction for: An inference workload maintains two always-on high-memory SageMaker endpoints for models invoked only during a brief nightly window. The decisive acceptance check is instead: The nightly workload still meets its completion target while cost reports show reduced idle serving hours.

The inference container accesses authorized model artifacts privately and an external egress test remains blocked. This would verify a different operation: Use approved private subnets, security groups and VPC endpoints where supported to restrict unnecessary network egress. That separate action addresses: A production inference container is placed in subnets with internet egress although its required datasets and dependencies are private AWS services. It cannot establish the required Harborline Logistics correction for: An inference workload maintains two always-on high-memory SageMaker endpoints for models invoked only during a brief nightly window. The decisive acceptance check is instead: The nightly workload still meets its completion target while cost reports show reduced idle serving hours.

Create a governed delayed-label collection path and compare production predictions with matched observed labels for model quality monitoring. It does not directly address the Northgate Telecom case: An inference workload maintains two always-on high-memory SageMaker endpoints for models invoked only during a brief nightly window. The relevant acceptance criterion for this case is: The nightly workload still meets its completion target while cost reports show reduced idle serving hours.

Use consistent random traffic allocation and analyze outcomes per experimental assignment and predefined business slices. It does not directly address the Northgate Telecom case: An inference workload maintains two always-on high-memory SageMaker endpoints for models invoked only during a brief nightly window. The relevant acceptance criterion for this case is: The nightly workload still meets its completion target while cost reports show reduced idle serving hours.

Monitor inference request schema and null-value rates against a versioned accepted contract before treating the endpoint as healthy. It does not directly address the Northgate Telecom case: An inference workload maintains two always-on high-memory SageMaker endpoints for models invoked only during a brief nightly window. The relevant acceptance criterion for this case is: The nightly workload still meets its completion target while cost reports show reduced idle serving hours.

A training role cannot approve its own model package and a designated release role can promote a reviewed version. This would verify a different operation: Separate training, model approval and production deployment identities with explicit role and approval boundaries. That separate action addresses: A machine learning pipeline uses a single IAM role to train models, approve registry packages and update production endpoints. It cannot establish the required Northgate Telecom correction for: An inference workload maintains two always-on high-memory SageMaker endpoints for models invoked only during a brief nightly window. The decisive acceptance check is instead: The nightly workload still meets its completion target while cost reports show reduced idle serving hours.

A representative request cohort produces traceable per-model score differences and unacceptable changes block promotion. This would verify a different operation: Run shadow or controlled A/B inference comparisons with logged model version and matched request IDs before full rollout. That separate action addresses: A retrained model is deployed without comparing its output distributions with the approved production model, causing unexplained score shifts. It cannot establish the required Northgate Telecom correction for: An inference workload maintains two always-on high-memory SageMaker endpoints for models invoked only during a brief nightly window. The decisive acceptance check is instead: The nightly workload still meets its completion target while cost reports show reduced idle serving hours.

A controlled prediction from the new variant appears in the correct monitoring dataset and the old variant is not mistaken for current. This would verify a different operation: Update monitored endpoint or variant configuration and baseline references as part of the versioned deployment process. That separate action addresses: A risk model is deployed with a new model version but its SageMaker Model Monitor schedule continues capturing only the old endpoint variant. It cannot establish the required Northgate Telecom correction for: An inference workload maintains two always-on high-memory SageMaker endpoints for models invoked only during a brief nightly window. The decisive acceptance check is instead: The nightly workload still meets its completion target while cost reports show reduced idle serving hours.