Hi there,
Though exact details are missing your requirements look quite clear and straightforward to implement. I would normally suggest to use a vm but the python pipeline looks like a perfect candidate to run on azure functions thus we wouldn't to deal with the hassle of running a full blown node. Also, since I don't know anything about the structure of data at this stage I'm not sure whether a relational db or a nosql solution would work better. If it's the former again I'm not sure if you have a specific choice, e.g sql server, cosmosdb, mysql, mariadb, ect.
Anyway, my first impression is that azure functions and sql server, just because you'd prefer it, is the stack we're going to use. Also, please let me know if you have a preference for python version to use. I expect this to take around 5 days to implement, thanks.