In Chapter 12, Data Exploration and Visualization, we explored a dataset by visualizing different features, using two packages—matplotlib and altair. The differences between those visualizations and a dashboard are twofold:

• The audience for the dashboard is meant to be wide, so it should be easily accessible via an internet browser. Visualizations are often made for self-consumption.
• Dashboards are meant to be frequently updated and, to some extent, interactive. Visualizations are often done on-the-spot, are static, and show only specific aspects of the data.

To a large extent, dashboards are full-blown projects requiring regular improvement and maintenance! However, as the demand for this kind of product is large, and the task is easily generalizable, there are plenty of solutions, tools, and services at our disposal. We can group them into three main categories:

• Third-party services, such as Tableau (there is still a running server behind it—it's just not yours to manage)
• Static—those without an active server behind them and all interactions happen on the client's machine
• Dynamic—those with an active server (or servers) behind them

Throughout this book, we focus on building stuff with code, rather than consuming, so we won't review services, of which there are dozens, now. If you're willing to build and support your own dashboards, it is totally possible! Let's study these in the upcoming sections.