Update: This was a nice experiment, but the code is not running efficiently anymore. R Markdown and various packages moved on. So this CV is outdates, too. Check my personal website, which uses simply markdown. It's simple and the beest.
My data-driven cv powered by Pagedown, a markdown setup.
I've followed nstrayer's hints and tweaked the style (that is, css). My CV embraces academic minimalism, especially with regard to the publications. There is no side-bar for publications, because there is no need (and no efficient code) to comment publications anyway. Check dd_cv.css for the improvements. There are also some custom edits in the cv.rmd file that improve the CV's accessability, this is where you should focus on.
I find this version much more suitable to a more conversative work environment, as opposed to the tech-savvy world of the original design.
Key to the datadriven cv is to have your own local or Google Drive cv data stored in a csv. Then, you simply need render the cv.rmd file. This is R markdown script, which is very simple code. Comment/outcomment with traditional html tags (<!-- ... ->>).
Please note that my code improvements are not super well documented. To understand the basics, check the original manuals.