Several Fully Funded PhD and Postdoctoral Positions in Foundation Models and Agentic AI for Physical Systems Saarland University in Germany

Saarland University is advertising several fully funded PhD and postdoctoral positions in the newly established Chair of Data-Driven Simulation and Analysis in Materials Science, led by Prof. Dr. Roland Aydin and jointly affiliated with the German Research Center for Artificial Intelligence (DFKI). The project theme is Foundation Models and Agentic AI for Physical Systems, combining modern AI with scientific computing, engineering, and materials research.

The research group works on LLMs and autonomous scientific agents that can reason about scientific problems, interact with simulation software, and support the design of complex physical systems. Current directions include AI for scientific computing, LLMs and scientific agents, AI for engineering design, multimodal scientific AI, alignment for scientific AI, and AI for materials discovery. The positions are embedded in Saarbrücken at Saarland University and DFKI, with strong ties to the Saarland Informatics Campus and international partners, providing access to a strong AI ecosystem and modern GPU/HPC infrastructure.

Successful candidates will pursue an independent research project and develop, train, and evaluate modern machine learning models such as graph neural networks, transformers, neural operators, diffusion models, and foundation-model-based systems. The work may combine machine learning with numerical simulation and scientific computing workflows. Researchers are also expected to publish at leading venues, present at international conferences, contribute to collaborative projects with EPFL and Hamburg University of Technology, and help support teaching and student supervision as appropriate.

Funding is fully provided under the German TV-L salary scale, pay grade E13, at 100% working time. The duration is 3 years for PhD positions and 3 years for postdoctoral positions, with a possible extension for postdocs. The job posting notes that E13 starts at approximately €57,112 gross per year before annual special payments, depending on prior experience and qualifications.

Applicants for the PhD track should have an above-average Master’s degree or equivalent in Computer Science, Computational Engineering, Materials Science, Mechanical Engineering, Physics, Applied Mathematics, or a closely related discipline. Postdoctoral applicants should hold a completed doctorate in one of these fields and have a demonstrated research record at the interface of AI and science. Strong Python skills, experience with PyTorch or JAX/TensorFlow, solid machine learning foundations, and excellent English (CEFR C1) are required; German is not required at the time of hiring. Helpful extras include publications, open-source contributions, and experience with diffusion models, foundation models, LLM fine-tuning, agentic AI systems, or scientific simulation software.

Applications must be submitted as a single PDF by 2026-08-15 to [email protected]. Applicants should use the correct reference number in the subject line: W2878 for PhD or W287G for PostDoc. The application should include a motivation letter, CV with publication list, transcripts or degree documents as appropriate, and two academic referees. Optional supporting links to GitHub, a personal website, or preprints may also be included.