Lasse Wrobel works as a research assistant for the BEI. He completed his bachelor's degree at the University of Leipzig in business informatics with a research focus on disruptive technologies. Since then, he has been investigating the influence of artificial intelligence and blockchain on companies in various industries. Currently, Lasse is studying for a Master's degree in Business Informatics.

Decentralized Identity – Secure Digital Identity Management?

Everyone is familiar with the following situation from everyday life: financial service providers or service providers (e.g., mobile network operators) offer services only for registered and verified users. The consequence: In order to be able to prove one’s own identity online, a new account must first be created using an e-mail address and a selected password. The process of creating and verifying different accounts results in a single user having many online identities and involves almost as many identity providers. Above all, the protection of one’s own data falls by the wayside in many cases. The advancement of blockchain/distributed ledger technology in recent years has given rise to a new approach to online identity processing and verification, Decentralized Identity. This post explores the concept as well as the underlying technology and highlights advantages over the traditional use of identity providers and user accounts.

Continue reading »

Digital Twins – The Merging of the Real and the Virtual World

The amount of data about real products, processes, and services has increased dramatically in recent years. This opens up new possibilities for planning, simulation and analysis. For this purpose, more and more companies use the concept of a digital twin. But what are digital twins and what potentials do they offer at the enterprise level, especially in the financial industry?

Continue reading »

Federated Learning – Efficient Machine Learning That Respects Privacy?

In the financial industry, customers expect high standards with regard to data protection and the integrity of their own data. Nevertheless, from the perspective of value creation, it is essential for banks to evaluate customer data using statistical methods and algorithms. Banks are thus caught in a conflict between maintaining data privacy and enforcing their own business model. To address this problem, the concept of “federated learning” has become established on the market in recent years, in which the data used for model training is always stored decentrally and the models are trained decentrally.

Continue reading »

“Data-centric AI” – A shift in the AI mindset?

The traditional approach to AI focuses on the process of training the model. The underlying data is often a secondary concern. This approach works particularly well for Internet corporations, as they have vast amounts of data and the capabilities to analyze it. In contrast, there is little potential for using AI in small businesses with this approach due to a lack of data. Therefore, it is worth taking a look at the data – moving away from model-centric to data-centric AI.

Continue reading »