Marlena Flüh, M.Sc.

Marlena Flüh, M.Sc.

Researcher

Ahornstraße 55, building E3 52074 Aachen Germany
9032
+49 241 80-21429
marlena.flueh@comsys.rwth-aachen.de

Short CV

Moin! Coming from the true north of Germany (somewhere north of Hamburg), I have been a PhD student at the Chair of Communication and Distributed Systems (COMSYS) since September 2025. Before that, I completed a vocational training as a software developer, studied Computer Science at TU Braunschweig (B.Sc.) and RWTH Aachen (M.Sc.), and gained research insights at DFKI, RWTH, and Fraunhofer FIT. In 2023, I was awarded a Femtec scholarship.

My research interests grew through the COMSYS Research Focus Class on Medical Data Privacy and my master’s thesis, where I worked on retrieval-augmented generation (RAG) over FAIR knowledge graphs for clinical research data. My PhD now focuses on the security and privacy of AI systems that process sensitive data, such as LLM-based agents and RAG systems in healthcare.

Projects

I was part of the interdisciplinary AI-MED project, which aimed to support diagnostics by analyzing radiology images using an agent-based AI system. Beyond image analysis, the system dynamically integrates additional patient information, while placing strong emphasis on data privacy, security, and compliance with regulations such as the GDPR.

Looking for a Thesis?

Are you interested in the security and privacy of LLM-based systems, especially in medical applications? If so, feel free to send me an email including your motivation, research interests, and a brief overview of relevant experience. I am also open to discussing your own research idea.

Interests
  • Security of multi-agent LLM systems
  • Traffic analysis of encrypted agent communication
  • Privacy in retrieval-augmented generation
  • Bias and fairness in synthetic medical data

Publications

Poster: Collaboration Fingerprinting Reveals What Adaptive Multi-Agent LLM Systems Process on the Wire. Proceedings of the 2026 ACM SIGSAC Conference on Computer and Communications Security. Presented at the 2026 ACM Conference on Computer and Communications Security (CCS '26), Nov 15 - Nov 19, 2026, The Hague, Netherlands. November 2026.
FAIR GraphRAG: A Retrieval-Augmented Generation Approach for Semantic Data Analysis. Proceedings of the 16th IEEE International Conference on Knowledge Graph (ICKG ‘25), Nov 13 - Nov 14, 2025, Limassol, Cyprus. November 2025.
ABEL: Artificial Buddy for Effective Learning. Proceedings of the 14th International Joint Conference on Knowledge Graph (IJCKG ‘25). Presented at the 15th International Joint Conference on Knowledge Graph, Oct 15 - Oct 17, 2025, Heraklion, Greece. October 2025.

Advised Theses

Please note that ongoing theses are only visible from within the RWTH network. If you are not affiliated with RWTH Aachen University and you are interested in our topics, please contact us.

A Topic in Privacy-Preserving Medical AI
Bachelor's Thesis, RWTH Aachen University, August 2026
Advisor: Marlena Flüh
Examiners: Klaus Wehrle, Daniel Truhn (UKA)
A Topic in Multi-Agent System Security
Bachelor's Thesis, RWTH Aachen University, June 2026
Advisor: Marlena Flüh
Examiners: Klaus Wehrle, Sandra Geisler (DSMA)
A Topic in Security & Privacy