Skip to content

Potential of the financial industry in the field of Data Science & Machine Learning

28. March 2022

Office Building, Building, Urban

The financial industry has been dealing with the digitalization of business processes and the analysis of data for many years. Banks want to take advantage of the enormous potential offered by technologies in the field of data science and machine learning. At the same time, however, they face technical, regulatory as well as organizational challenges. We highlight the current challenges with regard to Data Science and Machine Learning and present possible solutions.

Great potential in the use of Data Science & Machine Learning

Enormous efficiency and revenue gains can be realized in any business area through the use of data science and machine learning. Banks around the world have already begun to leverage the potential of data in various areas of their business, such as financial crime management, compliance management, product cross-selling, and sentiment analysis.

According to the study «AI Bank of the Future» by McKinsey & Company, AI can create additional value of up to USD 1 trillion annually in the global banking industry. AI and its subsectors are thus increasingly becoming an important factor in terms of future business success. Although banks recognize the benefits, implementing such projects is not easy.

Currently few projects implemented in the area of Data Science & ML

According to the study «How mature is AI adoption in financial services» by consulting firm PwC, less than a quarter of all AI initiatives within the DACH region are in a “live stage”. The study found that of the few ongoing AI projects, most are still in the early stages and make little contribution to overall automation and digitization strategies. Respondents cited the lack of available data, budget constraints and the lack of AI expertise within the company as the biggest challenges.

In addition, many of the respondents believe that companies must have a basic technical and organizational infrastructure in place for the effective use of AI. On the one hand, this requires a deep understanding of the processes and technologies related to data science and machine learning, and on the other, an established data strategy. Especially for highly regulated industries that work with sensitive data, ensuring transparency about AI processes is important for projects to be successful, legal and ethical. In the following, we address the challenges mentioned and present possible solutions.

Available data & budget constraint

Only 30% of the banks surveyed have a centralized approach to data collection and analysis, the majority of which, interestingly, are based in Switzerland. The data is mostly stored in data lakes, i.e. a large pool of raw data, and was collected primarily for regulatory purposes. While these Data Lakes are accessible, the data they contain is largely unstructured. This shortcoming is exacerbated by the low level of investment in AI, as IT resources often go to modernizing legacy systems rather than data science and ML projects. Instead of tackling extensive large-scale projects, we therefore recommend starting in small steps and implementing useful use cases using a Minimum Viable Product (MVP) approach.

AI Competence

One challenge that not only affects the financial industry is the shortage of Data Scientists. The reason for this is that the demand for data and statistics expertise in companies has increased sharply in recent years, while too few Data Scientists have been trained to meet the demand. Despite the growing interest and increase in data science training, experienced Data Scientists are still hard to find. Without this expertise, however, it becomes difficult to successfully implement Data Science projects. The transformation of business use cases into analytics questions (and vice versa) can be challenging.

Only 29% of AI projects in banks are developed internally, so the majority rely on external specialists. The cooperation can not only be aimed at the implementation of a project, but also at the training of employees. In addition to an experienced data science and software team, it requires a basic understanding of the subject matter by all project participants, from business analysts to departmental managers, in order to assess the possibilities and limits of the technologies and thus eliminate possible ambiguities as early as the definition of needs and goals.

Rely on long-standing data and software specialists

At Datahouse, data scientists, statisticians and software developers work closely together and jointly develop innovative solutions for you. From data acquisition and processing to the implementation of comprehensive data science and ML solutions, we support banks in their projects. We support you in all phases, from the definition of needs and goals to the implementation and maintenance of the solution. With our data packages, we offer you the optimal solution: flexible service packages according to scope, timeframe, and number of required data scientists and software developers. In this way, companies can find the right package for them to implement their project effectively.

One example of a data science project we implemented in the financial sector is the Portfolio Optimizer for Bank Vontobel. Our data scientists and software developers worked with Vontobel to develop rule-based models for optimizing the portfolio structure, which Asset Management uses to optimize the model and customer portfolios.

To support employees in your company in developing their data science competence, we offer our «Datahouse Academy – Data Science for Business» and give participants with different programming experience an introduction to business-relevant data science concepts. Are you planning a Data Science and/or Software project in your company and need professional advice and/or resources? Do not hesitate to contact us, we will be happy to advise you in a non-binding conversation.