AI-Powered Business Intelligence

Chapter 10. Bringing It All Together: Building an AI-Powered Customer Analytics Dashboard

You have come a long way in this book and (hopefully!) learned a lot of new things about how AI services can be deployed at various levels of the analytics stack. In this chapter, I will show you how these layers of analytics and AI services can be combined to provide a better experience for BI users and harness the potential of AI and BI by blending the two approaches. In fact, you should see that different AI services are not an either/or decision but that they can complement each other. For example, we can turn unstructured data (raw text) into structured data (sentiment scores table) so we can use it to do supervised learning (use sentiment scores to predict customer churn).

By the end of this chapter, you will be able to mix and match multiple AI services to develop even more powerful use cases. To keep programming effort to a minimum, we will use Azure Machine Learning Designer as a no-code tool to create advanced ML workflows that allow for more customization than the AutoML service you learned about in the previous chapters.

Problem Statement

In this scenario, we are part of the data analytics team of a telecommunications provider. The head of sales and marketing of the consumer division has initiated a project to look further into the topic of customer churn. As per the business’ definition, churn happens at the moment a customer cancels their contract, no matter the remaining duration ...