Explanation
"The decision of whether or not to file a suspicious transaction report (also known as a suspicious activity report or SAR in the United States) often involves weighing the aggravating and mitigating factors arising from the research conducted during the investigative process. Financial institutions should draft procedures that document the factors to consider when determining whether a suspicious transaction report (STR) is appropriate."

Use Azure Notebooks to develop a machine learning module and deploy it to a Linux device running Azure IoT Edge.
You can use IoT Edge modules to deploy code that implements your business logic directly to your IoT Edge devices.
References:
https://docs.microsoft.com/en-us/azure/iot-edge/tutorial-deploy-machine-learning

Explanation
https://www.bis.org/publ/bcbs85.pdf - page 12
Banks should gather sufficient information about their respondent banks to understand fully the nature of the respondent's business. Factors to consider include: information about the respondent bank's management, major business activities, where they are located and its money-laundering prevention and detection efforts; the purpose of the account; the identity of any third party entities that will use the correspondent banking services; and the condition of bank regulation and supervision in the respondent's country. Banks should only establish correspondent relationships with foreign banks that are effectively supervised by the relevant authorities. For their part, respondent banks should have effective customer acceptance and KYC policies.