About
I am an AI Research Engineer at Bosch Research in Renningen, developing next-generation agentic AI systems based on reinforcement learning for systems engineering — agents that interact with engineering artifacts, reason over goals and constraints, and improve through feedback, simulation, and optimization. My work combines hierarchical decision-making, model-based methods, and planning with LLM-based agents integrated with formal engineering tools.
Previously, I was a Senior Machine Learning Security Engineer at MATVIS GmbH in Tübingen, building the security foundation for safe generative AI deployment — including the AI Firewall and automated red-teaming for GenAI — and a PhD Visiting Researcher at Cambridge University, collaborating with Microsoft Cambridge on contextual LLM agents, their evaluation, and robustness, supervised by Prof. Dr. Per Ola Kristensson.
I hold a PhD in Computer Science from Saarland University, with a thesis on Continual Machine Learning and Model Adaptation for Personalized Systems, carried out as Deputy Head and Doctoral Researcher at the German Research Center for Artificial Intelligence (DFKI), where I continue as a consultant. At DFKI I have led a team of researchers and secured over 2 million Euros in research funding. My recent grants include €427,500 for Secure Language Models for Knowledge Management (SisWiss) and €100,000 for hybrid reinforcement and imitation learning (TeachTAM), both funded by the German Federal Ministry of Research (BMFTR, formerly BMBF).
My research covers Agentic AI & Reinforcement Learning, NLP & LLM Security & Evaluation, Incremental & Continual Learning, Imitation Learning, Multimodal Interaction & Interface Design, and Gesture Recognition & Computer Vision, applied to systems engineering, automotive, robotics, dialogue systems, and well-being domains, with partners including Bosch, Carl Zeiss, Microsoft Research, and BMW.