Context

Currently, dependency flow is managed by the Maestro++ service. This is a Service Fabric backed service which has been in place for several years. It is very stable but also very difficult to change as it's lacking in testing, missing a local developer workflow and other issues which make changes to the service very costly and risky.

For the reasons above, it doesn't sound like extending this service is possible with high velocity which might prevent fast iteration and timely delivery of the backflow dependency system.

Goal

Investigate whether we couldn't move onto a new, modern stack, that allows us to develop fast with high confidence. Ideally, build a proof of concept.
It should also be possible to move Maestro++'s logic into this new service and get rid of Service Fabric.

Notes

• Consider leveraging R9 libraries
• Consider the cost of servicing and maintenance such as S360