Summary

In this chapter, you learned how Power BI interacts with Microsoft AI services by default through Power BI Desktop and data flow features. You also learned that by using AutoML platforms, you can get around the licensing problem (PPU license or Premium capacity) that Power BI needs to interface with Microsoft AI services. You used both an on-premises AutoML solution (PyCaret) and Azure AutoML on the cloud to solve a binary classification problem. You also used Cognitive Services' Text Analytics to do some sentiment analysis directly using a Python SDK.

You've learned that enrichment via AI mostly happens in Power Query (which allows access to the internet), although you've seen a case where it may be convenient to use a machine learning model directly within a Python visual.

In the next chapter, you will see how to implement data exploration of your dataset in Power BI.