An ML engineer is optimizing a model that predicts customer behavior and needs to inspect input data and model predictions to identify patterns that might bias performance across demographics. Which solution provides this analysis?

SageMaker Clarify is the correct choice because it is specifically designed to detect bias in machine learning models and data. It can analyze training data for imbalances and evaluate model predictions for fairness across different demographic groups, directly addressing the need to inspect input data and model predictions for patterns that might bias performance. Amazon CloudWatch monitors infrastructure metrics, not data or model bias. AWS Glue DataBrew is for data preparation and cleansing, but it doesn't inherently analyze for bias patterns in the context of model performance. AWS Lambda functions are for serverless compute and automation, not for bias detection or fairness analysis.