ABSTRACT
Automation in data analysis generally boils down to writing software that replaces human operator in moving data around, transforming it into meaningful numbers and data visualizations, and delivering human-friendly outputs to stakeholders. While the ability to write code does not define a data scientist, it does act as a value multiplier. A good data scientist does not mind boring, mundane, and repetitive tasks, but understands what and why needs to be done, builds tools to have it done automatically, and moves on to new challenges. There are questions to consider before you automate something: if (it should be automated), when and how. When all is said and done, automation is just a meta-tool; its usefulness depends on the people wielding it.
