Machine Learning on Azure with automated predictions

Here, Azure Data Factory v2 (ADF) comes in handy.
ADF lets you connect to the on-premises database and write t-SQL queries to store data in the Azure Blob storage (blobs).
Predictive modelingOnce you have the required data stored in blob storage, you can use the Machine Learning Service Workspace (ML Services) platform that Azure provides for developing ML models.
Finally, you can save these predictions to Azure SQL Server Database or a blob as follows.
Connecting Databricks to Azure SQL Server DatabaseConnecting Databricks to Azure Blob StorageDashboardYou can connect the on-premises database with PowerBI once you start populating it with predictions.