Which solution will meet these requirements with the LEAST operational effort?
Use Amazon SageMaker to approve transactions only for products the company has sold in the past.
Use Amazon SageMaker to train a custom fraud detection model based on customer data.
Use the Amazon Fraud Detector prediction API to approve or deny any activities that Fraud Detector identifies as fraudulent.
Use the Amazon Fraud Detector prediction API to identify potentially fraudulent activities so the company can review the activities and reject fraudulent transactions.
Explanations:
Amazon SageMaker is a powerful tool for machine learning, but using it only for products the company has sold in the past does not directly address fraud detection. It limits the model’s scope and may result in false positives or missed fraud opportunities.
While SageMaker can be used to train a custom fraud detection model, this requires significant operational effort to gather and preprocess data, train the model, and monitor it continuously. The question specifies the need for minimal operational effort.
Amazon Fraud Detector is specifically designed to detect fraud with minimal setup. Using its prediction API to approve or deny activities that are flagged as fraudulent is the most efficient solution with the least operational effort.
This option would still require manual review, which adds operational effort. The goal is to minimize effort, and rejecting transactions automatically (Option C) is a better fit for this requirement.
I map out that the answer is:
Use the Amazon Fraud Detector prediction API to approve or deny any activities that Fraud Detector identifies as fraudulent.
I design that the answer is:
Use the Amazon Fraud Detector prediction API to approve or deny any activities that Fraud Detector identifies as fraudulent.