Q: 2
Scenario: A digital payments provider is experiencing fraudulent transactions, especially from new
accounts performing high-value payments. The existing batch model in SageMaker cannot flag activity
quickly enough. Goal: real-time fraud detection that can automatically assess and reject fraudulent
transactions at the moment of occurrence, requiring minimal operational effort.
Question- Which option satisfies these requirements?.
Options:
Options
Discussion
Why not C if you need custom fraud logic, or is managed API always preferred for lowest effort here?
C , had something like this in a mock and custom SageMaker model deployed to EC2 was accepted there.
C seems right for real-time inference if you need custom deployment, though it might not be as low effort as the others.
C deploying a SageMaker model to EC2 could do real-time inference if set up right. I saw a similar question in practice and they wanted custom code as a factor. Not sure but that's what I'd pick here.
Official AWS docs and practice exams both point to D.
D, that's exactly what Fraud Detector is for, lets you approve or deny on the spot with minimal ops.
Its D, Fraud Detector API does this out of the box for real-time checks, barely any setup needed.
Not B, D fits best for instant fraud checks as it's designed for real-time calls, so minimal ops work needed.
Its D, AWS Fraud Detector is actually built for these kinds of real-time scenarios, way less hassle than rolling your own model.
Be respectful. No spam.
