The use of streams is by no means new but it gives the analytics world a lot more flexibility than ever before. Even IT departments feel comfortable with streams as they can manage them but they still allow developers to create their magic in a live environment.

Of course, this approach isn't perfect and might not be useful to some types of industries. 

A big con of this approach is the fact that any person of a stream within the live environment can have access to live data. For data-sensitive industries such as healthcare, this approach is dangerous, and it will most probably be prohibited. After all, you wouldn't want a random consultant accessing your health history while developing dashboards, would you? 

But for many other industries, privacy isn't necessarily a problem and this approach is excellent as developers will be able to see precisely what dashboards will look like in a live environment and testers will be able to test the applications with real live and meaningful data. 

If data privacy is a problem, though, you can always talk to whoever is in charge of the data source and see whether it is possible to mask sensitive data such as names or addresses. By doing this, you still have the power of live data in your hands while protecting privacy.