Research

Research and publications.

Published research, a bachelor's thesis, and the academic side of the work.

Google Scholar Profile

Peer-Reviewed Publication

LLMs for Science: Usage for Code Generation and Data Analysis

Journal of Software Engineering and Applications (JSEP)·2024

Co-authored a peer-reviewed paper examining the use of Large Language Models in research and data analysis. The work has been cited over 130 times (35+ on Scopus), contributing to the academic discourse on responsible AI adoption in scientific workflows.

Impact: 130+ citations, 35+ on Scopus

Read Paper (DOI)View on Scholar
LLMsResearch MethodsCode GenerationData Analysis

Bachelor's Thesis

Echo: Mitigating Hallucination Potential in User Prompts Through AI-Guided Iterative Refinement

Technical University of Munich — School of Software Engineering and AI·2025

A novel shift-left approach to LLM hallucination mitigation. Introduced a user-sided hallucination taxonomy (Prompt Risk vs. Meta Risk), the Prompt Risk Density (PRD) metric for quantifying hallucination potential, and a multi-agent pipeline (Analyzer, Initiator, Conversation, Preparator) for iterative prompt refinement with human-in-the-loop validation.

Impact: Contribution to trustworthy AI in high-stakes domains (law, healthcare, finance)

Download PDFView Source
Hallucination MitigationTrustworthy AIPrompt AnalysisMulti-AgentNLP
Full Document

On research and practice

I find it useful to stay close to the research side. Not every paper translates to production, but understanding why something works makes building with it much more reliable. My academic work has mostly shaped how I think about failure modes and edge cases.