I am a Sr. Applied Scientist at Amazon in Berlin, where I specialize in developing efficient NLP and search solutions with a strong emphasis on system evaluation (metrics, synthetic data, auto-eval, analysis). Before joining Amazon, I completed my PhD at the UKP Lab, TU Darmstadt, where I regularly published in top academic venues (EMNLP, *ACL, AAAI, among others) and in 2020 became part of the senior staff. In 2011, I founded ARULABS UG, providing software consulting and development to direct clients and major advertising agencies (such as SYZYGY, Ogilvy, and ETECTURE)—building on my earlier work as a freelance web developer starting in 2008.
I am deeply interested in bringing cutting-edge NLP research into practical use—ranging from efficient LLM-based task automation and agents to automatic evaluation—and maintain a strong entrepreneurial drive, shaped by my background in founding and operating my own company. If you share these interests or have ideas for potential collaborations, please reach out!
During my PhD I frequently published in top academic venues including EMNLP, *ACL, AAAI, WWW, and NeurIPS. I also served as an area chair for the Efficient NLP track at EMNLP 2021–2023. In addition, I co-organized the Eval4NLP 2023 workshop on evaluation and comparison of NLP systems and the SustaiNLP 2022 workshop on simple and efficient NLP. Beyond these roles, I was a member of the ACL working group on Efficient NLP. Although I continue to conduct research at the forefront of NLP, my current role at Amazon emphasizes internal product development and real-world deployment over external academic publication.