Summary
Christian Merkwirth is an AI staff software engineer based in Krakow with two decades of ML/DL experience and a decade-plus track record of moving research into production across startups and Google. He has deep expertise in building and deploying deep learning systems (JAX, TensorFlow, PyTorch) and has led teams that shipped novel revenue-management and simulation systems, RL solutions, and production-grade model training pipelines. An early pioneer of graph neural networks—publishing one of the first molecular GNNs trained with SGD in 2005—he also adopted TPUs and custom trainer tooling years before mainstream adoption. Christian blends hands-on model engineering with product leadership, consulting across companies like Flyr, Nomagic and Maze while currently holding an AI staff role at Google and advising clients through nexocode. Notably, he pairs academic rigor (Max Planck research, university teaching) with pragmatic system-building, from C++ simulation cores to scalable DL trainers.
8 years of coding experience
22 years of employment as a software developer
Wilhelm-Filchner Gesamtschule
Biomathematics, Bioinformatics, and Computational Biology, Biomathematics, Bioinformatics, and Computational Biology at Max Planck Institute for Computer Science
Doctorate, Physics, Doctorate, Physics at The University of Göttingen
English, German, Polish, Russian