Calling web services in Power Query
Another way to interact with ML models in Power Query is to call web services. As you may already know, a ML model can be used to score many observations in batch mode against a trained model (the process described earlier). Another way to interact with a ML model is to expose it as a web service so that it can be called via REST APIs. You learned how to work with external APIs in Chapter 11, Calling External APIs to Enrich Your Data.
IMPORTANT NOTE
Keep in mind that you can’t consume external services via REST API calls from a Python or R visual because internet access is blocked for security reasons. Therefore, you can only consume these services in Power Query.
As an example, in this section, you’ll see how to invoke predictions from a released endpoint via Azure Machine Learning, and how to use the services exposed by the Azure Text Analytics feature of Cognitive Services. You could use some M code in...
