Predictive Analytics

This is why machine learners easily outperform humans: They are built for complex hyper-variable systems, which our brain is not. The latest predictors make use of a combination of prediction algorithms and sub-models and are able to cover short-term as well as long-term projections at the same time. Predictive models should not operate in isolation but rather always work in concert with simulation, decision, optimization or execution models. This is when they unfold their maximum ROI and business value. A careful overall design in the form of an integrated Algorithmic Target Operating Model (ATOM) is required to power the data driven business.

At the Steering Lab we were among the first to automate the search for predictive models. This approach solves most of the basic predictive problems quickly and efficiently and at the same time is almost maintenance-free, as the constant update and search for optimal predictive models and algorithms can be highly automated.

All this is part of the Steering Lab’s AAURUM framework. An extensive tool and template repository not unlike a data-modelling factory.

One of our predictive flagship solutions is the Global Business Radar suite. Business Radars based on Big Data are still very new concepts and very few companies have deployed these powerful predictive analytics machines. They scan our global socio-economy 24/7 and read and analyze hundreds of millions of documents relevant to industries and individual businesses. They detect trends, risks and opportunities and track events and their ripples through information space. Be that the spread of a pandemic such as the COVID-19 virus and the subsequent break-down of supply chains or the pick-up of the economy after such a crisis and the many darwinistic opportunities every crisis provides.