Data-driven problem solving and strategic operations beyond classical disciplinary boundaries.
Medicine teaches the art of making clear decisions under incomplete information. My analytical work forces me to methodically question those very decisions. I want to understand things, not just accept them. That's why I never look at problems in isolation, but try to grasp the mechanisms behind them, rather than merely treating symptoms.
I know clinical routines and their infrastructural hurdles firsthand, such as documentation that takes more time than the actual patient care. I approach such hurdles systematically, through data I analyze and processes I rethink, no matter where they arise. The methodology is a tool, not an end in itself. I don't look for the first available fix, but for approaches that respect real-world practice and hold up when applied.
What I offer isn't a fixed set of methods, but a specific way of thinking: I break complex issues down into their basic components, translate between disciplines, and develop concepts that work in practice. I apply this precisely where it has the most impact: in MedTech, strategic consulting, or at entirely new intersections.
Clinical pathophysiology, neuro-imaging, and multi-omics analysis. Deep understanding of ward workflows, real physician-patient interactions, as well as the technological limits and possibilities of therapy and diagnostics. Identification and leveraging of optimization potentials in medical data documentation.
Data-driven decision making, strategic controlling, logistics, and profound process and route optimization to increase efficiency in complex organizations.
Python, R, Java, JS, SQL. Local LLM deployment (Ollama, MLX, vLLM), PyTorch, LangChain, Qdrant (Vector DB). Automation & architecture via n8n, Docker, FastAPI, and WebSockets.
Automation & AI
Data Science & AI
Strategy & Logistics
Local e-commerce retailer operating across Europe (Camping equipment).
Development of a RAG pipeline for email communication with automated classification. Trained on >20,000 historical service tickets to automatically generate responses for simple inquiries.
Custom development of an Advanced Deep Synthesis Architecture.
A bipartite Python pipeline. Part A extracts, deduplicates, and semantically structures content from PDFs. Part B is a real-time FastAPI/WebSocket backend that transcribes documents and provides an interactive voice assistant via TTS.
European logistics company (as Junior Controller).
Analysis of revenue/expense streams, data-driven logistics and route planning in the Baltic Sea region, combined with direct client mediation and consulting.
For questions, offers or a non-binding consultation, I'm happy to help.
Send an email