INESC TEC developed intelligent system for risk analysis in the banking sector
The algorithms have been driving the banking business for a long time and it is no coincidence that Portuguese Manuela Velosa, one of the world's leading specialists in artificial intelligence, has led the innovation area of the largest American bank, JP Morgan since 2019. This is what took the French investment bank Natixis (from the universe of the Groupe Banque Populaire & Caisse d’Epargne, the second largest French banking group) to Porto.
It has been four years since Natixis arrived in search of Portuguese talent for a technological pole and even in January they announced that they will transfer 200 jobs in France to the offices of Porto, where the company has its' innovation hub already with close to a thousand employees.
Now they have partnered with the Institute for Systems and Computer Engineering, Technology and Science (INESC TEC), in Porto, to find a solution to correct financial series through algorithms. The analysis of risk in banking systems requires an exhaustive and repetitive work of evaluating millions of economic and financial indicators - a process that risk analysts are obliged to carry out to make decisions in credit operations or wealth management.
In this
partnership, the bank, and INESC TEC want to automate this process and limit
anomalies. "At Natixis, as in any bank, monitoring market risks involves
processing vast data sets, some of which may contain anomalies. In order to
improve our ability to detect and correct these anomalies, we have joined
together to INESC TEC to explore innovative solutions that can support the
daily tasks of analysts ", said Florent Soland, director of the
Informatics department at Natixis, in Portugal.
Until now "the data was presented to the analyst sequentially, on a website, and subsequently analyzed individually" and "whenever the analyst was faced with strange values, he was obliged to interrupt this analysis, in order to try to understand the origin of these values".
The new tools must now automate the detection of errors in the so-called vega values, that is, "volatile asset values, and understand the cause of these errors, in order to facilitate the work of analysts in the assessment of investment risk degrees", explained João Mendes Moreira, INESC TEC researcher and professor at the Faculty of Engineering of the University of Porto (FEUP), responsible for the project.
The researchers used data mining and machine learning techniques, which they applied to Natixis' risk assessment systems, in a protocol that is part of a set of projects promoted by a specialized INESC TEC structure, created to respond to the needs of companies in the financial sector, particularly with regard to its digitization processes.
INESC TEC has developed Artificial Intelligence solutions and other advanced technologies, applying them to financial processes, such as risk analysis, fraud detection, among others, in addition to standing out in other fields such as robotics or services (elaborated the Covid Stayaway app).
