Your team is architecting an application for an insurance company. The application will use a series of machine learning methods encapsulated in an API call to evaluate claims submitted by customers. Whenever possible, the claim is approved automatically but in some cases were the ML API is unable to determine approval, the claim is routed to a human for evaluation. Given this scenario, which of the following architectures would most aligned with current AWS best practices?
understand you should in general be using state machine/step functions with newer development, but since you are waiting for an external signal from the encapsulated API, shouldn’t you in fact in this case use simple workflow?
‘You should consider using AWS Step Functions for all your new applications, since it provides a more productive and agile approach to coordinating application components using visual workflows. If you require external signals to intervene in your processes, or you would like to launch child processes that return a result to a parent, then you should consider Amazon Simple Workflow Service (Amazon SWF).’
Appreciate any clarification/doc links
In this case, I can see where the API call might be considered an external signal, but what the FAQ means is something totally external to the flow of the process…like some manual approval for example. If, on the other hand, a manual review was right in the middle of the flow rather than one of the potential end results, I think AWS best practices would guide you toward SWF.