Machine Learning / AI expert in Garching bei München bei Max Planck Computing and Data Facility

The Max Planck Computing and Data Facility (MPCDF) is a competence centre of the Max Planck Society. It is engaged in the development and optimization of algorithms and applications for high performance computing as well as in the design and implementation of solutions for data-intensive projects in close collaboration with scientists of the Max Planck Institutes. The MPCDF operates Max Planck supercomputers, computer systems and data repositories for various Max Planck institutes, and maintains an up-to-date infrastructure for data management including long-term archival as one of the largest such academic sites within Europe.

The Max Planck Computing and Data Facility (MPCDF) is looking for a

Machine Learning / AI expert (m/f/d)
Job ID 09/2023

The successful candidate will be associated with the artificial intelligence (AI) and high-performance data-analytics (HPDA) division at the MPCDF (located at the research campus in Garching near Munich, Germany) and will work within the upcoming EU-Project “Magnetic Multiscale Modeling Suite” (MaMMoS) on machine learning and AI models to understand the magnetic properties of experimental and simulation data. Among others, the project aims at designing ML/AI models which can predict missing data (experimental and/or from simulations) in a complex parameter space and at correcting systematic errors of simulation results. A close collaboration with all the partners of the project, but especially with researchers of the Max-Planck Institute for the Structure and Dynamics of Matter (MPSD) will be key to the success of the project.

Key Responsibilities

• Finding and implementing corrections for systematic errors of simulation results
• Developing machine learning / AI models to predict missing data points (in the relevant parameter space) from either experimental measurements or simulation runs
• The interoperability and standards of data obtained from different experiments and different simulation code packages.
• The design and implementation of open-source software packages for all software tools developed within the project
• Documentation of the developed software and providing regularly project reports

Further details can be found on our project website.

Essential Skills and Qualifications

• PhD or equivalent qualifications in Physics, Materials Science, Computational Science, Engineering or a closely related field
• Experience and deep understanding of machine learning and AI methods
• High level of programming skills, including Python knowledge and its relevant libraries
• In-depth knowledge of at least one of the deep learning frameworks (Keras, Pytorch, Tensorflow, ...)
• Hands-on experience in feature-engineering for ML algorithms

Desired Skills and Qualifications

• Background in material science and especially experience and knowledge in the domain of magnetism and computational magnetism
• Experience in designing, implementing and optimizing machine and deep learning approaches
• Experience with classical machine learning regression algorithms, such as linear-regression, lasso-regression and regression trees
• Experience in analyzing the robustness of AI/ML solutions
• Experience working with high-performance computing systems
• Solid background in statistical analysis and the quantification of systematic errors and providing uncertainty estimations
• Knowledge about modern software design, software testing and experience in working on high-performance computing systems
• Knowledge of a programming language such as Fortran, C, or C++
• Experience in parallel programming with either OpenMP or MPI

Our offer

We offer the opportunity to work in an international and dynamic environment on cutting-edge, machine learning and AI projects at the interface of data from HPC simulations and experiments, in close collaboration with leading domain scientists of the Max-Planck society and their international partners. We also foster regularly research visits at the project partners institutions, but especially at the MPSD in Hamburg, Germany. Furthermore, we offer the opportunity to collaborate with specialists of major vendors of HPC systems on solutions for HPC-AI problems on existing and upcoming technologies.

The full-time position can start at the beginning of 2024 and is initially limited for a maximum period of four years. Salary will be based on experience according to the German TVöD.

Addit