Exam Bundle

A staged rollback restores the approved prior model version and representative requests return to baseline behavior. This would verify a different operation: Maintain versioned model artifacts and use a controlled deployment strategy with retained previous endpoint configuration. That separate action addresses: A production team deploys an updated fraud model by replacing the only model artifact, leaving no usable rollback option. It cannot establish the required Oakridge Public Services correction for: A cross-account promotion pipeline copies model artifacts but loses the KMS permissions required for the destination account to load them. The decisive acceptance check is instead: A destination test role retrieves the artifact and decrypts it while an unrelated account remains denied.

A clean model endpoint starts on the intended instance and no executable-format error appears in the startup logs. This would verify a different operation: Build and publish a runtime image for the target supported architecture and pin matching instance and container metadata. That separate action addresses: An inference platform chooses an EC2 instance family without ensuring that the container architecture matches its CPU instruction set. It cannot establish the required Oakridge Public Services correction for: A cross-account promotion pipeline copies model artifacts but loses the KMS permissions required for the destination account to load them. The decisive acceptance check is instead: A destination test role retrieves the artifact and decrypts it while an unrelated account remains denied.

A known request traverses the versioned preprocessing and prediction stages and produces the expected output after deployment. This would verify a different operation: Use a supported SageMaker inference pipeline or reviewed multi-container design with compatible input and output contracts. That separate action addresses: An inference service chains preprocessing and prediction in application code that cannot be versioned with the SageMaker deployed model. It cannot establish the required Oakridge Public Services correction for: A cross-account promotion pipeline copies model artifacts but loses the KMS permissions required for the destination account to load them. The decisive acceptance check is instead: A destination test role retrieves the artifact and decrypts it while an unrelated account remains denied.

Use SageMaker Batch Transform or an appropriate scheduled batch inference workflow instead of an always-on real-time endpoint. It does not directly address the Meadowbrook Finance case: A team declares its pipeline finished when model training succeeds, though deployment smoke tests and runtime metrics are never observed. The relevant acceptance criterion for this case is: A controlled deployment must pass endpoint health, representative inference and alarm checks to reach the completed state.

Build and publish a runtime image for the target supported architecture and pin matching instance and container metadata. It does not directly address the Meadowbrook Finance case: A team declares its pipeline finished when model training succeeds, though deployment smoke tests and runtime metrics are never observed. The relevant acceptance criterion for this case is: A controlled deployment must pass endpoint health, representative inference and alarm checks to reach the completed state.

Use a supported SageMaker inference pipeline or reviewed multi-container design with compatible input and output contracts. It does not directly address the Meadowbrook Finance case: A team declares its pipeline finished when model training succeeds, though deployment smoke tests and runtime metrics are never observed. The relevant acceptance criterion for this case is: A controlled deployment must pass endpoint health, representative inference and alarm checks to reach the completed state.

An oversized test input is queued, processed and retrieved asynchronously without the client's synchronous timeout. This would verify a different operation: Choose SageMaker asynchronous inference or a supported job-based flow with input-output object locations and completion handling. That separate action addresses: An image analysis API accepts infrequent large payloads with long processing times but times out when called synchronously through a normal endpoint. It cannot establish the required Meadowbrook Finance correction for: A team declares its pipeline finished when model training succeeds, though deployment smoke tests and runtime metrics are never observed. The decisive acceptance check is instead: A controlled deployment must pass endpoint health, representative inference and alarm checks to reach the completed state.

A replay of transient spikes avoids oscillation while sustained load still triggers the required scale-out action. This would verify a different operation: Tune target tracking, cooldown and minimum capacity from observed workload periodicity rather than setting a zero-tolerance threshold. That separate action addresses: An auto-scaled inference endpoint scales out during every brief request spike and incurs unnecessary instance churn and warm-up overhead. It cannot establish the required Meadowbrook Finance correction for: A team declares its pipeline finished when model training succeeds, though deployment smoke tests and runtime metrics are never observed. The decisive acceptance check is instead: A controlled deployment must pass endpoint health, representative inference and alarm checks to reach the completed state.

Requests from representative client Regions meet the permitted total latency without violating data-residency controls. This would verify a different operation: Benchmark end-to-end inference latency and select an approved Region or edge design aligned to residency and client location. That separate action addresses: A low-latency image model deployed to a generic instance experiences unnecessary network delay because request payloads transit a distant AWS Region. It cannot establish the required Meadowbrook Finance correction for: A team declares its pipeline finished when model training succeeds, though deployment smoke tests and runtime metrics are never observed. The decisive acceptance check is instead: A controlled deployment must pass endpoint health, representative inference and alarm checks to reach the completed state.

Deploy an appropriately scaled low-latency real-time inference endpoint and route transaction requests through the authorized path. It does not directly address the Redstone Aviation case: A team declares its pipeline finished when model training succeeds, though deployment smoke tests and runtime metrics are never observed. The relevant acceptance criterion for this case is: A controlled deployment must pass endpoint health, representative inference and alarm checks to reach the completed state.

Implement the required model loading, health and invocation handlers for the supported SageMaker custom container interface. It does not directly address the Redstone Aviation case: A team declares its pipeline finished when model training succeeds, though deployment smoke tests and runtime metrics are never observed. The relevant acceptance criterion for this case is: A controlled deployment must pass endpoint health, representative inference and alarm checks to reach the completed state.

Compare Lambda, SageMaker asynchronous inference or container hosting based on payload size, memory and duration constraints. It does not directly address the Redstone Aviation case: A team declares its pipeline finished when model training succeeds, though deployment smoke tests and runtime metrics are never observed. The relevant acceptance criterion for this case is: A controlled deployment must pass endpoint health, representative inference and alarm checks to reach the completed state.

The model container loads from authorized private resources and an unapproved subnet remains unable to reach the endpoint. This would verify a different operation: Configure the supported SageMaker VPC subnet, security group and endpoint-access rules for required private dependencies. That separate action addresses: A SageMaker endpoint runs inside private subnets but the security groups omit the intended dependency paths and container startup fails. It cannot establish the required Redstone Aviation correction for: A team declares its pipeline finished when model training succeeds, though deployment smoke tests and runtime metrics are never observed. The decisive acceptance check is instead: A controlled deployment must pass endpoint health, representative inference and alarm checks to reach the completed state.

A replay of transient spikes avoids oscillation while sustained load still triggers the required scale-out action. This would verify a different operation: Tune target tracking, cooldown and minimum capacity from observed workload periodicity rather than setting a zero-tolerance threshold. That separate action addresses: An auto-scaled inference endpoint scales out during every brief request spike and incurs unnecessary instance churn and warm-up overhead. It cannot establish the required Redstone Aviation correction for: A team declares its pipeline finished when model training succeeds, though deployment smoke tests and runtime metrics are never observed. The decisive acceptance check is instead: A controlled deployment must pass endpoint health, representative inference and alarm checks to reach the completed state.

The endpoint becomes healthy and returns a valid prediction for a documented JSON request through the SageMaker invocation API. This would verify a different operation: Implement the required model loading, health and invocation handlers for the supported SageMaker custom container interface. That separate action addresses: A platform team deploys a custom PyTorch inference container without the correct SageMaker-compatible serving endpoint contract. It cannot establish the required Redstone Aviation correction for: A team declares its pipeline finished when model training succeeds, though deployment smoke tests and runtime metrics are never observed. The decisive acceptance check is instead: A controlled deployment must pass endpoint health, representative inference and alarm checks to reach the completed state.

Redact or omit sensitive input fields from logs and emit privacy-safe structured metadata with governed retention. It does not directly address the Highland Education case: A customer churn model starts receiving features with a distribution that differs sharply from training, but the endpoint tracks only CPU and availability. The relevant acceptance criterion for this case is: A controlled shifted feature batch triggers the intended drift alert while a reference batch remains within its baseline threshold.

Instrument authorized distributed traces or structured per-stage timings, using AWS X-Ray where supported, to isolate bottlenecks. It does not directly address the Highland Education case: A customer churn model starts receiving features with a distribution that differs sharply from training, but the endpoint tracks only CPU and availability. The relevant acceptance criterion for this case is: A controlled shifted feature batch triggers the intended drift alert while a reference batch remains within its baseline threshold.

Configure precise cross-account package, S3 and KMS key permissions for the trusted production consumer role. It does not directly address the Highland Education case: A customer churn model starts receiving features with a distribution that differs sharply from training, but the endpoint tracks only CPU and availability. The relevant acceptance criterion for this case is: A controlled shifted feature batch triggers the intended drift alert while a reference batch remains within its baseline threshold.

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 Highland Education correction for: A customer churn model starts receiving features with a distribution that differs sharply from training, but the endpoint tracks only CPU and availability. The decisive acceptance check is instead: A controlled shifted feature batch triggers the intended drift alert while a reference batch remains within its baseline threshold.

A load test that violates the p99 target triggers an alarm even though mean latency remains acceptable. This would verify a different operation: Track high-percentile ModelLatency and relevant invocation or backlog metrics with thresholds based on the service-level objective. That separate action addresses: An endpoint's p99 latency increases during bursts but CloudWatch averages look stable and no tail-latency alarm is configured. It cannot establish the required Highland Education correction for: A customer churn model starts receiving features with a distribution that differs sharply from training, but the endpoint tracks only CPU and availability. The decisive acceptance check is instead: A controlled shifted feature batch triggers the intended drift alert while a reference batch remains within its baseline threshold.

The development user can create approved experiments in the correct domain but receives AccessDenied for protected production operations. This would verify a different operation: Apply identity-based domain and resource permission boundaries with segregated workspace roles and least privilege. That separate action addresses: A SageMaker Studio user can create experiments in a production domain even though their intended access is restricted to development. It cannot establish the required Highland Education correction for: A customer churn model starts receiving features with a distribution that differs sharply from training, but the endpoint tracks only CPU and availability. The decisive acceptance check is instead: A controlled shifted feature batch triggers the intended drift alert while a reference batch remains within its baseline threshold.

Pin a governed immutable model artifact version and validate checksums or signatures through the approved release process. It does not directly address the Maplewood Schools case: A customer churn model starts receiving features with a distribution that differs sharply from training, but the endpoint tracks only CPU and availability. The relevant acceptance criterion for this case is: A controlled shifted feature batch triggers the intended drift alert while a reference batch remains within its baseline threshold.

Enable appropriate tuning-job completion limits and early termination policies where supported for the selected metric. It does not directly address the Maplewood Schools case: A customer churn model starts receiving features with a distribution that differs sharply from training, but the endpoint tracks only CPU and availability. The relevant acceptance criterion for this case is: A controlled shifted feature batch triggers the intended drift alert while a reference batch remains within its baseline threshold.

Use AWS Secrets Manager or an approved secret provider with a scoped runtime role and suppress secret values from pipeline logs. It does not directly address the Maplewood Schools case: A customer churn model starts receiving features with a distribution that differs sharply from training, but the endpoint tracks only CPU and availability. The relevant acceptance criterion for this case is: A controlled shifted feature batch triggers the intended drift alert while a reference batch remains within its baseline threshold.

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 Maplewood Schools correction for: A customer churn model starts receiving features with a distribution that differs sharply from training, but the endpoint tracks only CPU and availability. The decisive acceptance check is instead: A controlled shifted feature batch triggers the intended drift alert while a reference batch remains within its baseline threshold.

The runtime writes to the intended log stream but cannot modify an unrelated service's logging resources. This would verify a different operation: Scope CloudWatch log actions to required resource ARNs and separate log configuration permissions from ordinary invocation needs. That separate action addresses: A model deployment can write arbitrary CloudWatch Logs groups because its inference IAM role includes broad administrative logging permissions. It cannot establish the required Maplewood Schools correction for: A customer churn model starts receiving features with a distribution that differs sharply from training, but the endpoint tracks only CPU and availability. The decisive acceptance check is instead: A controlled shifted feature batch triggers the intended drift alert while a reference batch remains within its baseline threshold.

The benchmark identifies the accelerator bottleneck and the selected corrective capacity change reduces inference failures. This would verify a different operation: Monitor the appropriate GPU utilization, memory and throughput metrics for the selected hosting architecture. That separate action addresses: A CloudWatch dashboard claims the inference host is idle because CPU is low, yet GPU memory saturation causes inference throttling. It cannot establish the required Maplewood Schools correction for: A customer churn model starts receiving features with a distribution that differs sharply from training, but the endpoint tracks only CPU and availability. The decisive acceptance check is instead: A controlled shifted feature batch triggers the intended drift alert while a reference batch remains within its baseline threshold.

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 Lakeside Research case: A production classifier's false-negative rate rises over several months, but predictions are logged without collecting any eventual ground-truth outcomes. The relevant acceptance criterion for this case is: A reference set with known labels produces a measurable recall trend and an agreed quality alert fires when the threshold is breached.

Pin a governed immutable model artifact version and validate checksums or signatures through the approved release process. It does not directly address the Lakeside Research case: A production classifier's false-negative rate rises over several months, but predictions are logged without collecting any eventual ground-truth outcomes. The relevant acceptance criterion for this case is: A reference set with known labels produces a measurable recall trend and an agreed quality alert fires when the threshold is breached.

Instrument authorized distributed traces or structured per-stage timings, using AWS X-Ray where supported, to isolate bottlenecks. It does not directly address the Lakeside Research case: A production classifier's false-negative rate rises over several months, but predictions are logged without collecting any eventual ground-truth outcomes. The relevant acceptance criterion for this case is: A reference set with known labels produces a measurable recall trend and an agreed quality alert fires when the threshold is breached.

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 Lakeside Research correction for: A production classifier's false-negative rate rises over several months, but predictions are logged without collecting any eventual ground-truth outcomes. The decisive acceptance check is instead: A reference set with known labels produces a measurable recall trend and an agreed quality alert fires when the threshold is breached.

A test model role reads its permitted subset and is denied access to unrelated original customer attributes. This would verify a different operation: Publish an approved data subset or governed view and grant read access only to the required machine learning project roles. That separate action addresses: A team shares a full customer dataset across projects by broadening S3 bucket access instead of providing a minimally scoped training extract. It cannot establish the required Lakeside Research correction for: A production classifier's false-negative rate rises over several months, but predictions are logged without collecting any eventual ground-truth outcomes. The decisive acceptance check is instead: A reference set with known labels produces a measurable recall trend and an agreed quality alert fires when the threshold is breached.

The production role loads the shared artifact and a separate untrusted account still receives an access denial. This would verify a different operation: Configure precise cross-account package, S3 and KMS key permissions for the trusted production consumer role. That separate action addresses: A cross-account model registry package is shared with a production account but the referenced encrypted S3 model artifact cannot be decrypted there. It cannot establish the required Lakeside Research correction for: A production classifier's false-negative rate rises over several months, but predictions are logged without collecting any eventual ground-truth outcomes. The decisive acceptance check is instead: A reference set with known labels produces a measurable recall trend and an agreed quality alert fires when the threshold is breached.