requirements completed university studies (master/diploma) in the field of chemistry, chemical engineering, environmental chemistry, data sciences, geosciences or related field possess a solid background in geochemical processes (exempli gratia, sorption, speciation, radionuclide transport) complement the chemical expertise with some experience in machine‑learning or data‑analytics tools high‑level programming skills (python, r, julia) to build, test, and optimize models of geochemical systems interest in large‑scale computational simulations (exempli gratia, reactive transport, sorption processes) Capability to work in a structured, solution‑oriented manner, demonstrating analytical thinking and a strong commitment to project goals Motivation to work collaboratively in an interdisciplinary and international team-oriented environment Excellent communication skills in English

Tasks You will help modelling complex geochemical systems, which are typically limited by extremely high computational demands. To break this bottleneck and cut simulation time by orders of magnitude, you will design and implement surrogate models that learn the behavior of full‑physics codes using modern machine‑learning techniques. These surrogates will be tuned for rapid, uncertainty‑aware predictions and integrated into decision‑support tools for deep geological repositories of nuclear waste—one of the most pressing challenges facing modern societies. Specifically, the tasks are: Identify state‑of‑the‑art machine‑learning (ML) methods that can be applied to geochemical systems in geological contexts Assess these methods for traceability, robustness, and physicochemical correctness Implement and adapt the most promising algorithms to model radionuclide migration in crystalline host rocks Execute proof‑of‑concept ML simulations and perform a risk analysis of the resulting model outputs Present your scientific results at conferences, workshops, and seminars, and publish the work in peer‑reviewed journals Collaborate with project partners at CASUS (HZDR), TU Bergakademie Freiberg, and TU Darmstadt