The "data technologies" pyramid is an easy to understand representation of the various technologies we have worked with. In general, the higher you go in the pyramid, the higher the level of automation and the higher the skills required to achieve the intended results. Data reporting is a well understood discipline and most platforms support it well, its easy to use but it's value is limited to when the report was run; the next level up, data analytics provides other dimensions, like time and personas, and can be used to automate a lot of the manual data reporting tasks from the lower level; analytics typically includes a way to forecast performance based on the data it has collected in the past but there is no learning capability. Machine learning and natural language processing are examples of using extensive processing capabilities to enhance the level of automation in a software-based system. Ultimately, one end-goal of this system is 100% automation of business processes, so the system "understands" the insights, automatically makes a "decision" and takes the optimum "action" to drive optimum performance.