Companies and institutions of all kinds need the new predictive analysis tools more than ever. The first steps in this discipline were dominated by descriptive and diagnostic tools that helped shape what, over time, came to be called business intelligence. But gradually this technology, central to the fourth and fifth industrial revolutions, became oriented towards the future.
In general, technicians have used several of these systems at the same time. This practice means that professionals in this field have advanced training and are constantly updating their knowledge. In return, they enjoy high salaries and excellent working conditions, at least in developed countries.
As the experts recall, predictive analysis is a complementary field to ordinary management aimed at forecasting, in an automated way, what could happen in a sector or industry —with a greater probability— through the detailed and in-depth study of patterns and trends, always based on large volumes of past and present data.
The traditional thing was that a small and select group of data scientists were in charge of this task, as complex as it was influential. The processes they led could require weeks, or even months, of experimentation. Hypotheses were explored and prototypes were validated to find the best model, the one that added the most value to the interested company or organization.
However, as Gartner research director Carlie Idoine points out, the landscape “is changing dramatically”. The simplification of the software and greater versatility is allowing less specialized users to obtain a significant benefit. Even the names are now easier to understand and retain: data science, that is, data science; and machine learning, that is, automatic learning.
Artificial intelligence has made it possible for a sales manager to interact with an algorithm that scores their customers, actual and potential, to make better decisions. Or that a marketing manager knows how to improve the volume of clicks that your web page receives. And that the finance team finds a way to reduce the volume of fraud or defaults…