Skip to content

ML Solution Monitoring, Maintenance, and Security Questions

Practice questions for ML Solution Monitoring, Maintenance, and Security topic in AWS Certified Machine Learning Engineer - Associate. 48 questions covering this domain.

48 questions12 easy25 medium11 hard
Q1
medium

A security team needs to deliver a long-term record of AWS account activity to Amazon S3, with optional delivery to CloudWatch Logs and Amazon EventBr...

Q2
easy

An operations team needs to interactively search and analyze application logs in Amazon CloudWatch Logs during incident response. Which feature should...

Q3
medium

A company needs AWS to protect its root encryption keys, enforce access through key policies, and keep the keys from leaving the service unencrypted. ...

Q4
medium

A team wants to compare production inference data against statistics computed from the training dataset and be alerted when the live data violates exp...

Q5
medium

Traffic to a SageMaker-hosted model rises during business hours and drops overnight. The team wants AWS to add and remove instances automatically so i...

Q6
easy

A security engineer needs a searchable, downloadable, immutable record of the past 90 days of management events in an AWS Region at no CloudTrail char...

Q7
medium

A company wants one SageMaker service that can monitor data quality drift, model quality drift, bias drift, and feature attribution drift for deployed...

Q8
hard

An operations team repeatedly runs CloudWatch Logs Insights queries against very large log groups and wants better query performance with reduced scan...

Q9
medium

A team wants to standardize how it queries log data in CloudWatch and chooses among Logs Insights QL, OpenSearch PPL, and OpenSearch SQL. Which featur...

Q10
hard

A team launches SageMaker Studio in a custom Amazon VPC and wants Model Monitor to function correctly without public internet access. What additional ...

Q11
easy

An ML application currently stores database passwords in source code. The team wants a managed AWS service to store and rotate those secrets securely....

Q12
medium

A platform team wants secrets encrypted by default and also wants optional automatic rotation for those secrets. Which statement about AWS Secrets Man...

Q13
hard

An ML platform team wants centralized governance and audit of cross-account SageMaker activity, including who invoked which endpoints and when. Which ...

Q14
medium

A team wants to control SageMaker endpoint access so only specific principals can invoke a deployed model. Which mechanism should be used?

Q15
hard

A regulated team needs to keep SageMaker endpoint inference traffic from leaving the AWS network and reaching the public internet, even when calling o...

Q16
easy

Which AWS service is the recommended store for ML training artifacts and model packages, providing 11 9s of durability and integration with SageMaker?

Q17
medium

Which Amazon CloudWatch capability lets ML teams collect custom metrics like inference latency or throughput emitted from SageMaker endpoints?

Q18
hard

A SageMaker endpoint accidentally returns predictions for production traffic that includes PII. Compliance asks the team to discover and remediate sen...

Q19
medium

A team wants to capture inference requests and responses from a SageMaker real-time endpoint for monitoring without modifying the inference code. Whic...

Q20
medium

An ML team wants to encrypt SageMaker notebook instance and training EBS volumes using a customer-managed key. Which AWS service should they use?

Sign in to see all 48 questions

Create a free account to browse all questions — completely free during our launch phase.