Thesis Topics

Tools that leverage GenAI technology have been shown to enhance user productivity in content creation tasks. Noy and Zhang (2023) found that writers using OpenAI’s ChatGPT produced content more quickly while maintaining improved overall quality. Similarly, GitHub's Copilot, marketed by Microsoft, claims to increase software development speed by 55.8% (Peng et al., 2023). 

GitHub Copilot is powered by Codex, a large language model trained on publicly available GitHub repositories (Chen et al., 2021). It assists software developers in two primary ways: through a conversational interface, similar to ChatGPT, which answers software engineering queries, and through advanced code completion mechanisms that streamline coding tasks and also generate complete parts of software code. 

While tools like GitHub Copilot, ChatGPT, and other GenAI models are being rapidly adopted by software developers worldwide, their impact on due diligence (DD) and mergers and acquisitions (M&A) remains unclear. Solutions such as Sema (Sema Software) enable organizations to scan codebases for AI-generated code to assess potential risks. However, the specific risks and benefits of GenAI-generated code in the context of DD and M&A, particularly its influence on firm valuation, are not yet well understood. 

This master's thesis examines the impact of AI-generated code on the M&A process, specifically its implications during due diligence. It aims to evaluate the potential risks, opportunities, and overall effects of AI-generated code in acquisitions. The key questions explored include: How can AI-generated code positively influence due diligence? And, in what cases might its use pose risks that could jeopardize an acquisition?

Methodology:
Case-study

Language of the thesis: English

Getting started with literature:

• Andriole, S. (2007). Mining for Digital Gold: Technology Due Diligence for CIOs.
  Communications of the Association for Information Systems, 20, pp-pp.
  https://doi.org/10.17705/1CAIS.02024
• Noy, S., Zhang, W., 2023. Experimental evidence on the productivity effects of
  generative artificial intelligence. Science 381, 187–192.
  https://doi.org/10.1126/science.adh2586
• Peng, S., Kalliamvakou, E., Cihon, P., Demirer, M., 2023. The Impact of AI on Developer
  Productivity: Evidence from GitHub Copilot.
• Chen, M., Tworek, J., Jun, H., ..., 2021. Evaluating Large Language Models Trained on
  Code. https://doi.org/10.48550/arXiv.2107.03374
• Bain & Company, 2024. Harnessing Generative AI in Private Equity,
  www.bain.com/insights/harnessing-generative-ai-global-private-equity-report-
  2024/

Supervisor: David Rochholz

Enterprise Architecture (EA) principles serve as fundamental guidelines that influence the development and governance of an organization’s IT landscape. However, the effectiveness of these principles depends on their structured dissemination, ensuring that strategic, high-level guidelines translate into actionable, operational principles. This cascading process must not only function in a top-down manner but also integrate bottom-up feedback mechanisms to maintain alignment with real-world requirements.

This master's thesis explores how the cascading of Enterprise Architecture Principles (EAPs) affects their effectiveness in guiding architectural decisions and aligning IT systems with business strategy. By analyzing the interactions between strategic, tactical, and operational levels, the study aims to identify best practices for ensuring that EA remains a practical and adaptive framework rather than a detached theoretical construct.

Methodology:
Qualitative or Quantitative 

Language of the thesis: English or German

Getting started with literature:

• Haki, K., & Legner, C. (2021). The mechanics of enterprise architecture principles. Journal of the Association for Information Systems, 22(5), p. 1334–1375.

Supervisor: Elias Grewe

Enterprise Architecture (EA) is a critical framework for aligning IT strategy with business objectives. However, its effectiveness is not solely determined by formal structures and cascading principles but is also significantly shaped by organizational culture. Culture influences how EA principles are adopted, interpreted, and applied within an organization, potentially reducing the need for rigid cascading structures by embedding EA principles into everyday practices and decision-making.

This master's thesis investigates how organizational culture affects the implementation and evolution of Enterprise Architecture. The study explores the relationship between cultural factors, such as shared values, norms, and communication styles, and the governance mechanisms of EA frameworks. It examines whether a strong organizational culture can replace or reduce the necessity for formal cascading mechanisms in EA adoption.

Key Research Questions:

• To what extent can a strong culture substitute formal EA governance structures?
• What mechanisms within an organization's culture facilitate or hinder EA implementation?

Literature to get started: 

Aier, S. (2014). The role of organizational culture for grounding, management, guidance and effectiveness of enterprise architecture principles. Information Systems and e-Busi-ness Management, 12, 43–70.

Methodology: Quantitative or Qualitative

Language: English or German

Supervisor: Elias Grewe

Traditional Enterprise Architecture (EA) frameworks are built around structured design principles, static rules, and hierarchical decision-making processes. However, the rise of AI-driven architectures, often referred to as Software 2.0, is reshaping these paradigms. Software 2.0 introduces machine-learned models that replace deterministic systems leading to dynamic, data-driven, and self-evolving systems. This shift challenges established EA frameworks, requiring new methodologies for governance, adaptability, and integration.

This master's thesis explores how Software 2.0 and AI-driven architecture components influence Enterprise Systems. It investigates the necessary adaptations, the implications for decision-making, and the governance mechanisms required to ensure alignment between AI-driven systems and traditional system components.

Key Research Questions:

• How does Software 2.0 redefine traditional Enterprise Architecture principles and practices?
• What challenges and opportunities arise from integrating AI-driven components into EA frameworks?
     

Methodology: Qualitative

Language of the Thesis: English or German

Getting Started with Literature:

• Dilhara, M., Ketkar, A., & Dig, D. (2021). Understanding software-2.0: A study of machine learning library usage and evolution. ACM Transactions on Software Engineering and Methodology (TOSEM), 30(4), 1-42.
     

Supervisor: Elias Grewe

This design research will explore strategies for optimizing crawling workflows for empirical research.

Your Tasks:

• Analyze existing crawling architectures and identify optimization opportunities concerning empirical research questions
• Develop a data engineering architecture for empirical research
• Implement and evaluate efficiency improvements in real-world crawling scenarios
• Ensure compliance with ethical and legal constraints in web data collection

Your Profile:

• Strong programming skills (Python, Java, or similar)
• Knowledge of web protocols
• Interest in empirical research and data engineering
• Ability to work independently

Supervisor: Elias Grewe

As systems grow more complex and interconnected, managing and understanding their underlying architecture becomes increasingly challenging. Digital representations of technical architectures, such as digital twins, are essential for addressing this complexity. This design research explores strategies for creating digital representations, facilitating better decision-making, and optimizing enterprise technological infrastructures.

Your Tasks:

• Develop a framework for discovering and representing technical architectures.
• Showcase how this framework can support decision-making processes and enhance the efficiency and scalability of enterprise infrastructures.

Entry Literature:

• Weller, M. (2024, September). Automated Synchronization of Enterprise Architecture Models with Deployment Models. In Proceedings of the ACM/IEEE 27th International Conference on Model Driven Engineering Languages and Systems (pp. 136-141).

Your Profile:

• Strong programming skills (Python, Java, or similar)
• Strong data engineering skills
• Ability to work independently

Supervisor: Elias Grewe

As enterprise systems become increasingly complex, optimizing and maintaining architectural integrity becomes critical. A significant portion of valuable architectural knowledge is often tacit, residing in the minds of technical architects rather than in formalized, codified documents. This research focuses on developing a framework for collaborative optimization of enterprise architecture through machine learning methods. The goal is to explore strategies to learn from past human-driven changes, progressively improving its ability to suggest architectural enhancements.

Your Tasks:

• Develop a methodology that uses machine learning to analyze historical enterprise architecture changes and learn from past decisions made by human architects.
• Integrate rule-based systems to apply predefined logic for identifying inefficiencies, risks, and opportunities for improvement within the architecture.
• Enable the system to provide data-driven suggestions for future architectural changes based on patterns identified in past human decisions.

Entry Literature:
Borozanov, V., Hacks, S., & Silva, N. (2019). Using machine learning techniques for evaluating the similarity of enterprise architecture models. In Advanced Information Systems Engineering: 31st International Conference, CAiSE 2019, Rome, Italy, June 3–7, 2019, Proceedings 31 (pp. 563-578). Springer International Publishing.

Your Profile:

• Strong programming skills (Python, Java, or similar).
• Some experience with machine learning algorithms and logic programming (e.g. Prolog)
• Strong problem-solving and analytical abilities.
• Ability to work independently

Supervisor: Elias Grewe