Most companies reach a chaotic stage in the evolution of their business by cobbling together ad-hoc systems to process data into advanced analytic output using scripts and non-specialized tools. These rudimentary production systems often deliver initial customer satisfaction, but the costs of managing data this way grow exponentially, eat up the profits that should come with the economies of scale, and act as an impediment to business growth. They are far from the data factory model and high-performance data factory architecture.