MagicOrange Incorporates New Feature: MACHINE LEARNING

Machine learning seems to be the ideal technology to compliment a cost transparency solution like MagicOrange. Due to its ability to learn constantly, to recognize similar data sets over time and, as new forms of data are input, to automatically match these new data inputs to existing cost categories. This automation can accelerate existing processes from weeks or months to mere days or even minutes.

A cloud-based cost transparency solution, MagicOrange is hosted in Microsoft Azure, enabling it to easily integrate with Azure Machine Learning, Microsoft’s own cloud-based analytics service for building, training and utilizing machine learning models. In this way, MagicOrange are not only able to make use of the vast library of best-in-class algorithms in Azure, but also have the ability to build other models using widely accepted R or Python libraries and queries.

One of the first applications of Machine Learning that MagicOrange utilized was to send consumption and financial data through a proprietary algorithm, within Azure Machine Learning. This allow the generation of highly granular amounts and relationships between cost model inputs and every activity, service or business unit in the model, regardless of where in the allocation or value chain it sits. This new insight was delivered via the MagicOrange ‘Front to Back Report’. This report enables a holistic view of cost flow – from source to final destination, allowing the relationship between the source General Ledger Financial and the ultimate end recipients to be understood, and vice versa, and so further drive understanding and open up even more cost optimization opportunities.

Azure Machine Learning neatly integrates with the Azure stack – which is the MagicOrange cloud database service – enabling the various Azure Machine Learning models to be exposed to the MagicOrange application. The beauty of this is that by having the Azure Machine Learning capability readily available and tightly integrated with existing cost transparency solution, a wide range of cost modelling approaches can be adopted.

In other words, everything from predictive financial modelling to clustering of similar services, and anomaly detection to advanced industry benchmarking can be implemented in low code or no code models. This, in turn, opens up exciting possibilities for MagicOrange and for the manner in which our customers can consume their cost transparency solutions.

Contact us today to learn more about our MagicOrange and its Machine Learning feature and how it can assist your organization.