AIF-C01 Machine Learning Operations (MLOps) 3

Question 1

Which AWS service can be used to create event-driven model retraining pipelines by reacting to new data uploaded to S3?

Accepted Answer

Amazon EventBridge

Answer

AWS Batch

Answer

AWS Step Functions

Answer

Amazon SQS

Question 2

What does 'training-serving skew' mean in ML operations?

Accepted Answer

When the features or data preprocessing used during training differ from those used during inference

Answer

When a model trains faster than it can be deployed

Answer

When the model's latency is higher at serving time than during training

Answer

When different teams use different SageMaker instance types for the same model

Question 3

Which SageMaker Model Monitor schedule type detects changes in data quality of live inference input data?

Accepted Answer

Data Quality Monitor

Answer

Model Quality Monitor

Answer

Bias Drift Monitor

Answer

Explainability Monitor

Question 4

In MLOps, what is the purpose of a model approval gate in a deployment pipeline?

Accepted Answer

To require a human or automated review before a trained model is promoted to production

Answer

To automatically approve all models that exceed a minimum accuracy threshold

Answer

To restrict which AWS accounts can access model artifacts in S3

Answer

To enforce encryption of model artifacts at rest

Question 5

What is the benefit of using Amazon SageMaker Pipelines over AWS Step Functions for ML workflows?

Accepted Answer

SageMaker Pipelines has native integrations for ML-specific steps like training, processing, and model registration

Answer

SageMaker Pipelines supports more AWS services than Step Functions

Answer

SageMaker Pipelines is free while Step Functions charges per state transition

Answer

Step Functions cannot orchestrate parallel tasks whereas SageMaker Pipelines can

Question 6

What is ML lineage tracking and why is it important?

Accepted Answer

Recording the complete chain of artifacts, steps, and parameters that produced a specific model

Answer

Tracking the geographic origin of training datasets for compliance

Answer

Monitoring the network lineage between SageMaker endpoints and S3 buckets

Answer

Auditing IAM role changes that affect SageMaker training jobs

Question 7

Which statement best describes the 'continuous delivery' (CD) component specific to MLOps?

Accepted Answer

Automatically deploying approved models to a staging or production environment after passing validation gates

Answer

Automatically pushing every trained model directly to the production endpoint

Answer

Delivering new features to the ML platform infrastructure without downtime

Answer

Continuously streaming new training data to a running SageMaker training job