AI Engineer · Cologne, Germany

I build pragmatic LLM systems that hold up in production.

I work on enterprise GenAI applications, evaluation, and structured extraction pipelines. My focus is turning emerging AI capabilities into systems teams can actually trust and use.

Selected signals

  • 100k+ PDF pages processed per month
  • 3,500+ retail stores touched by deployed AI
  • 150+ colleagues upskilled in agentic tooling

About

From apprenticeship to applied AI.

I started my career through an apprenticeship at Deutsche Telekom and later moved into business informatics, research, and AI engineering. What stayed consistent across all stages was a strong bias toward practical problem solving, fast learning, and getting useful systems into the hands of real users.

Today, I work at the intersection of enterprise constraints and frontier model capabilities. I enjoy building systems where LLMs are only part of the story and the harder work is making them fit real workflows, evaluation standards, and organizational reality.

Work

Selected projects and experience.

REWE Group · Current

Enterprise GenAI systems for production use

Built an LLM-powered ETL pipeline processing more than 100,000 PDF pages per month and co-developed a hybrid-search GenAI endpoint used across more than 3,500 markets.

Forward deployed work

Label extraction agent for retail master data

Embedded with a business team to ship an end-to-end workflow from incoming assets to structured output and manual approval. The result reduced work from roughly 30 minutes to 50 seconds per product at about 95% accuracy.

Enablement

Internal consulting and AI adoption

Supported business units in shaping AI roadmaps and upskilled more than 150 colleagues in agentic engineering workflows with tools such as Claude Code, Codex, and GitHub Copilot.

2025 — today

Junior ML Engineer, REWE Group

Enterprise GenAI applications, evaluation, and LLM systems.

2024 — 2025

Working Student, Research & Innovation Engineering

Evaluated agentic systems and built proof-of-concept architectures for wider enterprise AI adoption.

2022 — 2025

M.Sc. Business Informatics, TU Wien

Focus on LLM-based systems, evaluation, and context-aware emotion analysis.

2018 — 2022

B.Sc. Business Informatics, WWU Münster

Built the academic foundation for moving from business systems into applied software and AI work.

Research

Context-aware emotion analysis with LLMs.

Publication · ACM CHI EA 2025

Context over Categories: Implementing the Theory of Constructed Emotion with LLM-Guided User Analysis

A peer-reviewed publication on a context-aware approach to emotion analysis based on Lisa Feldman Barrett's theory of constructed emotion.

Read publication

Master Thesis · TU Wien

Context over categories: implementing the theory of constructed emotion

Operationalized a psychological theory into an LLM pipeline by introducing the “context sphere” for nuanced, context-dependent emotion analysis.

Read thesis