AI-Powered Business Intelligence

Chapter 5. AI-Powered Descriptive Analytics

In this chapter, we will explore two use cases that illustrate how AI can help us perform descriptive analytics faster and provide a more intuitive and seamless way to interact with large datasets—even for nontechnical people. We will also see how the natural language capabilities of modern BI tools (in this example, Power BI) can take away some of the mundane tasks from us.

Use Case: Querying Data with Natural Language

The typical analytical thinking process of a business user often starts with a simple question, such as one of these:

• How were my sales last month?

• Did we sell more this year than last year?

• What’s the top selling product?

These questions are mostly descriptive in nature. Users first need to understand the status quo before they can dive deeper into analysis or even anticipate future events.

The classic way to solve this problem would be for an analyst to create a set of static reports for the most common questions for specific business units or departments. However, it’s extremely difficult for business analysts or BI designers to anticipate all of the most common questions by using static reports or visualizations. Depending on which problem you look at and which people you ask, the exact area of interest can be fundamentally different. In organizations, this usually leads to long, exhaustive meetings with different stakeholders discussing which charts to display in a dashboard.

This is where self-service BI ...