Welcome back. In this course, we've covered a bunch of topics like dealing with dirty data in module one, creating and maintaining tables in module two. In module three, we got lots of practice developing complicated queries for real-world applications. You'll need all of these skills in this module to go through this case study. This module is all about using your SQL skills to set up a basic AB testing system which is really an important application of data science because we build stuff all the time, and we always hypothesized that what we build is better, better for the user, better for the bottom line et cetera, but we can't prove or disprove that claim without testing it both ways. That's where AB testing comes in. You need causal inference to prove or disprove your hypothesis. AB testing is an application of a kind of statistics called hypothesis testing. The idea is that you start with a hypothesis like, this new improvement to a website will cause some metric to increase. Typically, this metric is tied directly to some business metrics like revenue, but it might also make sense to look at metrics that are more explanatory. So, you can prove or disprove a hypothesis about how user behavior changed. By looking at multiple metrics, you can interpret the results holistically to gain insight about the effects of the change. As the first data scientist at a company, it might be your job to setup AB testing from scratch, and this module can be your guide. On a more established data science team, you may not be building all of these pieces but you might be responsible for creating new metrics related to the specific product area you're working on. Even if your main focus as a data scientist is on machine learning, and algorithm development, AB testing on real users is an important final step in the algorithm development process, where you verify that the new iteration is an improvement on the existing version. So, you'll need to understand how it works. Even if you're not expecting to work on experiments, this is a good example of an SQL project that's bigger in scope than developing a single query. There are of course more complicated variations of AB testing that we won't cover in this class, such as comparing more than two variants or looking at interaction effects between multiple concurrent experiments. But this module and the final project for the course should test your SQL skills and give you the basic experience you need to learn anything more complicated that comes your way.