So last but not least, I did mention machine learning and there's no discussion of machine learning that you can't have without explaining the difference between what does a Data Scientist do versus what a Data Engineer does versus a Data Analyst? So let's jump into a quick discussion of that before your lab. All right. So Data Analysts, you are the target audience for this course. So we do have a Data Engineering course after this one. So if these concepts and data really just excites you and you're like, man, I really want to build machine learning models, get through this course and then do an amazing job at learning a lot of these core concepts because all these data concepts build on each other, one after another after another. You can't build machine learning models without having first a fundamental understanding of processing your data and understanding how to sample it and understanding what is dirty data, right? So, largely a merge of technologies between the three, with kind of the key differences between the rules, right? Data Analysts, like those of you taking this course, you'll be writing SQL inside a BigQuery Web UI and visualizing that data. Data Scientists would be using something like online collaborative notebook, like a IPython notebook, and you can have more of a statistics background, and you'll be building these machine learning models, right? And you could be using SQL, but in addition, in other languages like R and Python. Data Engineers, by contrast, you're building these amazing pipelines of torrents of petabytes of data, right, that are coming in to these systems. So you have logs data that's being streamed from an online video game and you need to store that data somewhere, and then need to pipe that to your data analysts and BigQuery for that to be available for analysis as well. All right. I mentioned that online kind of gaming analytics. We'll close with one of these examples before summarizing and getting you guys into your lab. So, highlighted in red is where your Data Analysts could come into play, right? So you have two different types of data. You have streaming events coming off of video game online. You could also have these massive logs of data that's loaded in batches. And that's loaded into that staging area that we talked about, Google Cloud Storage through that bucket resource. You're going to be creating one of those in just a few modules. And then, as we mentioned, Data Engineers could use something like Cloud Dataflow to build a pipeline that you can then pipe in massive amounts of data into your warehouse like Google BigQuery to perform these ad hoc queries. And then further right of that, for your reports, you could use something like Google Data Studio to analyze, explore and visualize, and present that information. Or if you're a Data Scientist, you can just plug Cloud Datalab as a layer on top of BigQuery and invoke those queries to preprocess your data to build something like cool machine learning models to see, "Hey, if we needed to optimize this level, or too many people are failing on this boss, we predict that maybe if we adjusted this parameter, let's see what would happen there". All right, lets walkthrough some of those tools in the toolbox. And again, this is just to get you an introduction to a lot of these big data tools for analysis. And if you're so interested in a lot of these technologies, definitely feel free to just keep taking these courses. To Summarize what we covered so far. We looked at the lifecycle of Data Analyst tasks and mapped each of those tasks to the right tools to use on Google Cloud platform. We compared data roles and toolsets used by Data Analysts, Data Scientists, and Data Engineers. While this course is primarily targeted towards Data analysts, it will provide a clear ramp up into more advanced tools and topics that are covered in greater depth than other courses like Data Engineering. Next up, let's continue our foray into BigQuery, by practicing dataset exploration.
