Hi, I had to do something similar having hit a functional issue with Azure SQL.
I followed this document
https://learn.microsoft.com/en-us/azure/virtual-machines/windows/sql/virtual-machines-windows-portal-sql-create-failover-cluster?sa=X&ved=2ahUKEwjI8YX4odznAhWD34UKHe7nBoAQ9QF6BAgIEAI
It’s a little out of date but the approach seem to work well.
On the backup side you’ve got a couple of options;
Azure backup for SQL server
SQL backup on board the cluster but send the files to a mounted Azure file store
We’ve used Azure backup for SQL but also pulled a copy of the nodes themselves via ASR due to the criticality of the dataset..
Hope this helps a little.