Knowing how to manage data properly can give companies competitive advantages. Jean-Bernard Nobs and his team are the big data specialists at PostFinance. In our interview with him, he explains exactly what his work entails.
Big data – big solutions: how our data experts work
What impact can data experts actually have within a financial institution? Jean-Bernard Nobs works at PostFinance in the Data Science & Artificial Intelligence team. We ask him six questions about big data.
Data mining, data analytics or data science: which of these terms best describes what you do?
Data science, of course. After all, data mining and data analytics are just sub-areas of data-related work. Data science, however, is what we ourselves do: we give our companies and customers an advantage by using data to create future-proof solutions to various problems. But it is important to stress that processing and presenting data is not seen as an end in itself. Our data-related work, which involves using tools from various data disciplines, such as deep learning and visualization, is comprehensive and creates added value.
How exactly does data science help improve banking?
Data science can have both an internal and external impact. The work we do helps improve the efficiency of our operational banking processes through automation, on one hand, and, on the other, data also helps us understand our customers better. Data science allows us to tailor the measures we take.
On data protection: with the General Data Protection Regulation (GDPR), data protection has become more restrictive. What does this mean for your work?
This clearly creates new obstacles. On the other hand, however, it also presents an opportunity to come up with new solutions that are not immediately apparent. For customers willing to share data, there is scope to create solutions that really can help them.
How does a data science project actually work?
To start off with, as we have already mentioned, there is always some sort of problem, for instance: “How can we reduce the number of manual steps involved in identifying fraudulent activity?” Once the problem at hand has been set out, we are confident of our chances of success and we know whether we will see a positive return on investment, we start looking for data, compiling it in such a way that we have a reasonable pool of it. If the quality of the data is good enough, we attempt to process it (e.g. using computer-assisted visualization techniques) so that we can analyse it and gain some insights. Next, depending on the project, we use tools, such as machine learning, to resolve the problem.
What role does data science play at PostFinance?
An increasingly significant one. This is something that our development over the past few years alone shows us: since 2014, our team has increased from two to ten people. Like nearly all large companies, we are still not able to tap into the full potential of our data world at the moment because it is not that easy to compile all this data due to the fact our systems have increased in size. The data is still distributed too heavily in silos, but we are working on a central database – as well as on proving the added value of data science at every available opportunity. This means we often have a lot of persuading to do, especially when the data paints a completely different picture to what our own experience suggests.
About Jean-Bernard Nobs
Jean-Bernard Nobs has been working at PostFinance in the Data Science & Artificial Intelligence team since 2015. He studied maths and physics at the Saint-Michel school in Fribourg, graduated with a Bachelor and Master’s in Life Science and Biotechnology from the Swiss Federal Institute of Technology in Lausanne, and also obtained a PhD in Bioengineering and Biotechnology. Why he enjoys working as a data science specialist at PostFinance in his own words: “As a data science specialist, you establish correlations between business and data, and do your bit to make your company successful. Our work is also really exciting: the issues we deal with are always different, and analysis presents brand-new issues for us to tackle.”