The development of Agentic AI has given rise to a new form of automation that goes beyond traditional, rule-based RPA. These intelligent agents can make decisions autonomously, act in a context-sensitive manner and control complex tasks independently.

The aim of this thesis is to investigate the potential applications of Agentic AI in the financial sector. Two case studies will be used to illustrate and evaluate specific usage scenarios. 
 

A data-driven culture describes a corporate culture in which data-based decisions are systematically promoted, supported and integrated into everyday work. It is a key prerequisite for fully exploiting the potential of business analytics and successfully implementing data-driven innovations. This is not just about technical solutions, but above all about changing mindsets, skills and leadership styles.

The aim of this thesis is to identify success factors and obstacles in establishing a data-oriented corporate culture.
 

With the increasing use and analysis of large amounts of data, the risk of data breaches and security vulnerabilities is also growing. Privacy Enhancing Technologies (PETs) offer innovative technical approaches to protect personal data while enabling its use for analytical purposes. These include methods such as differential privacy, homomorphic encryption, secure multi-party computation, and data anonymisation.

The aim of this thesis is to present key PETs and evaluate their potential for use in data-driven companies.
 

Algorithm aversion describes the phenomenon that people often trust algorithmic decisions less than human ones – even when algorithms deliver objectively better results.

The aim of this paper is to analyse, based on existing empirical studies, whether and to what extent individual risk propensity is related to algorithm aversion. 

Algorithm aversion describes the phenomenon that people often trust algorithmic decisions less than human ones – even when algorithms deliver objectively better results. The causes of this behaviour are manifold and only partially understood so far.

The aim of this thesis is to systematically analyse, based on current research literature, which factors trigger or reinforce algorithm aversion and how these factors interact. 
 

Visualisation standards in dashboards play a central role in the effective communication of data within companies. Well-designed dashboards enable decision-makers to quickly grasp complex information and make informed decisions. However, it is often unclear which forms of visualisation, design principles and design rules are considered "best practices".

The aim of this thesis is to examine key visualisation principles and standards. Among other things, it will examine issues such as information architecture, colour choice, diagram types and interaction options in order to derive basic recommendations for the design of dashboards.