Is Babel’s Data Tower Compatible With Centralized ML Solutions?
Blogs from ODSC SpeakersConferencesModelingEurope 2019posted by ODSC Community November 14, 2019 ODSC Community
Does data have a common language? Can there be ML solutions for everyone?
Data structure and its content reflect the technological domain where the data comes from. The available technology is a relevant aspect to be considered. Professionals usually create databases by using the tools which are available in the digital market.
Therefore, most of the data is collected and aggregated with cutting-edge digital technologies and it is affected by improvements in data management. Users are another important factor because the effectiveness of technology depends on the way it is utilized.
Let us consider data coming from several Customer Relationship Management (CRM) tables of a financial institution. These tables include different data from a variety of countries, but they functionally represent the same kind of customer information. That being said, each country might have its own data language. This is due to different people, market features, management decisions and adopted technologies.
Consider a case in which many CRM professionals meet to solve a business problem impacting their daily work. These professionals might come from different countries. As CRM experts, they need to identify who within their population is potentially interested in buying a specific product, or in need of a tailored service in order to predict a decrease/increase in their business’ volume. They are aware that even the smallest change in customer habits may be relevant for volume trends. When the reason is an attrition between business and consumer, CRM professionals should investigate its causes and put in place mitigation actions, where necessary.
Here we are: a common goal. Everything seems clear. But the business environment is even more complex and dynamic. Many aspects should be considered. Let’s highlight two of them: goals peculiarities and data languages. The former refers to the different approach that each country uses to accomplish its objectives. Indeed, the way a business grows is typically related to the characteristics of its local market, economic system and so much more. In the same way, marketing campaigns are specifically tailored for each market. As data language, the fundamental differences between data structures across countries is intended.
There is often the need to find a common path, by finding a sort of ’Rosetta Stone’ of data that could sift through all the differences to find a common language.
The identified ‘Rosetta Stone’ process is a Data Science pipeline compatible with all the needs of local CRM teams. Even though the goals might be different, it is possible to find a common skeleton by removing unnecessary details. All the goals could be addressed with a common approach. This allows for the creation of a shared solution pipeline, i.e. a unique software.
An efficient software configuration is achieved only with the right balance between generalization and specialization. The former to enforce maintenance, the latter to match all the business needs.
Editor’s note: Ivan and Federica are speakers for ODSC Europe next week! Check out their talk, “A Centralized Customer Relationship Management Approach For Banking,” there!
Ivan Luciano Danesi has been working in UniCredit Services S.C.p.A. since July 2014 as a Data Scientist. His activities have been mainly focused on Risk Management and on Customer Relationship Management within the Big Data framework. He achieved a Ph.D. in Statistics at the University of Padua, in which he carried research activity in collaboration with Università di Trieste (Trieste) and CASS Business School (London). He is Adjunct Professor and research collaborator at Università Cattolica del Sacro Cuore (Milan).
Federica Perugini has been working in UniCredit Services S.C.p.A. since late 2015 as a Data Scientist. Her activities have been mainly focused on Business-To-Customer applications and on Customer Relationship Management within the Big Data framework. She achieved a M.Sc. in Applied Mathematics at La Sapienza University in Rome. She is a teaching assistant at M.Sc. in Data Science at Università Cattolica del Sacro Cuore (Milan).