BDC integration with Azure Machine Learning Service as external compute option
Some customers have datasets they can't share to the public cloud and they also don't have good compute options on-premise. The marriage of BDC as a compute option to Azure Machine Learning Service would be a good way to distribute training of models that must be trained on-prem and still take advantage of the superior UX in AML. In particular, the Jupyter and AI User Interface on AML are much stronger than Azure Data Studio. Getting out of running tools on workstation and using browser based AI tools is what customers want.