Preface

Welcome! If you are interested in this book, you’re most likely familiar with Power BI, machine learning (ML), and OpenAI. Over the years, Power BI has evolved from a data visualization tool into a suite of user-friendly, end-to-end software as a service (SaaS) tools for data and analytics. I began writing this book with the goal of teaching Power BI professionals about the ML tool built into Power BI. Rather than write a technical manual, I decided to embrace the tradition of Microsoft’s popular Power BI Dashboard in a Day course (found at https://aka.ms/diad) by writing the book as an end-to-end journey, starting with raw data and ending with ML, all within the SaaS Power BI toolset.

During the course of writing the book, an amazing new technology arrived on the scene called OpenAI. OpenAI can generate and summarize human language in amazing ways. The use case for this book was a perfect fit for adding OpenAI as a capstone to the journey.

This book will take you on a data adventure starting with real raw data from the Federal Aviation Authority (FAA), reviewing requirements that mimic a real-world project, cleansing and curating the data using Power BI, making predictions using Power BI ML, and then integrating OpenAI into the use case. You can recreate the entire end-to-end solution by referencing the Packt GitHub site (https://github.com/PacktPublishing/Unleashing-Your-Data-with-Power-BI-Machine-Learning-and-OpenAI/) as you read the book.

Business Intelligence (BI), ML, and OpenAI use data in different ways requiring different data modeling techniques and preparation. In my experience, most Power BI professionals think about data differently from ML and AI professionals. When BI professionals first branch into ML, these differences can cause ML projects to fail. Through the example of a real data story, this book attempts to teach those differences in the context of a use case with similar challenges and requirements to those that you may face in the real world. The overarching theme is the intersection of these skill sets for real-world projects that seamlessly incorporate BI, ML, AI, and OpenAI.

If you are looking for a technical manual about Power BI ML or OpenAI, this book is not for you. This book will walk you through a hero’s journey that builds up to ML and OpenAI as a capstone to the project. At the end of this book, beyond understanding how to use Power BI ML and OpenAI, you will understand how to think about and understand data projects in ways that can incorporate ML and OpenAI. Even if the tools in Power BI evolve to be different from the time this book was written, you should be able to apply these learned lessons to new tools and future challenges.

I also want to briefly discuss SaaS ML tools in this preface. I’ve often heard experienced ML professionals urge caution with regard to SaaS ML tools. I agree that ML as a discipline requires a different mindset and unique skillset from many other data tools. Many factors can lead to ML models returning misleading or biased results. ML projects that need to be highly accurate, or that could have harmful outcomes when wrong, should be handled by ML professionals using advanced ML tools.

That being said, a SaaS tool such as Power BI ML still has a powerful place with the right audience. Power BI professionals interested in learning about ML can skill up quickly by using Power BI ML. Rapid feature discovery, simple predictive use cases, and ad hoc hypothesis testing can all be achieved with a low bar to entry using Power BI ML. The ML models you will build in this book are intended to spark your interest in the subject, not provide a comprehensive course on building proper ML models. By the end of this book, a Power BI professional will understand the basics of why they might use ML, how data needs to be modeled for ML, and how ML can be used in the workflow of a data project. Hopefully, some of you are inspired to learn more about ML and graduate to more advanced ML tools and courses.

Regarding OpenAI, the final two chapters provide use cases for OpenAI that add value to the hands-on workshop with the affiliated GitHub workshop. Real FAA data is used to generate new descriptions and summarize events in your Power BI solution. The intent of this book is not for you to become OpenAI or ML experts, but rather to understand the intersection of BI, ML, AI, and OpenAI. It is my belief that as enterprise SaaS tools such as Power BI become easier to use, the intersection of these skills and tools is the future of our profession.