Karlsruhe Institute of Technology (KIT) is recruiting a Postdoctoral Researcher in the Scientific Computing Center to work on the BMFTR-funded MorphoSphere project, developing scalable AI methods for analysing ultra-high-resolution 3D imaging datasets from synchrotron and neutron facilities. The role focuses on building distributed deep learning algorithms for petabyte-scale image processing using HPC systems and multi-GPU architectures. Responsibilities include designing scalable segmentation and landmark detection models, optimizing workflows within the HeAT framework, implementing GPU-accelerated solutions, and validating methods on real scientific data. The researcher will collaborate with domain scientists, contribute to research software engineering best practices, and publish results in international peer-reviewed venues