Traineeship
Jülich Supercomputing Centre (JSC), Forschungszentrum Jülich GmbH
Research Institute / National Laboratory
6 months
01/10/2026
Enabling AI agents to bridge Quantum Computing and HPC
HPCTRAIN-OFFER-202601-120
AI & Machine Learning, Quantum Computing & Hybrid HPC-QC, Quantum Computing, High-Performance Computing (HPC), Scientific Software Engineering, Heterogeneous Workflows
Internship on AI-Driven Hybrid Quantum computing and HPC workflow
The trainee will join the scientific and technical team at the Jülich Supercomputing Centre (JSC), Forschungszentrum Jülich, working under the supervision of a scientific mentor and collaborating with experts in High-Performance Computing (HPC), AI, and quantum computing. The trainee will take on the role as a research software developer and AI workflow engineer.
Bachelor's degree completed, Master's student, PhD candidate, The ideal candidate should have a completed Master's degree or be a highly motivated candidate with a completed Bachelor's degree (PhD candidates and early-career researchers are also welcome). Relevant fields of study include Computer Science, Computational Science & Engineering, Physics, Mathematics, Data Science, or a related technical discipline. Essential Skills: - Strong programming proficiency in Python. - Familiarity with Linux/UNIX command-line environments, shell scripting, and Git version control. - Eagerness to bridge AI technologies with high-performance computing infrastructure. Desirable Skills: - Exposure to modern AI-driven coding tools, autonomous agents, or the Model Context Protocol (MCP). - Experience with containerization technologies (Docker, repo2docker, Apptainer/Singularity). - Familiarity with interactive computing environments (Jupyter/JupyterLab). - Basic knowledge of quantum computing concepts or scientific workflow tools.
Domain-specific application development, Machine learning workflows on HPC, Scientific software engineering, Workflow & job management, Quantum Computing Workflows, Reproducible Scientific workflows
Git / GitLab, Linux / Bash, Python, Singularity / Apptainer, Slurm, FastMCP, Qisktit, myQLM, Jupyter-AI
View the HPCTRAIN call for trainees call textThe traineeship focuses on the practical development and optimization of software infrastructure for AI-driven hybrid workflows, intersecting AI tools, classical High-Performance Computing (HPC), and quantum computing resources. The trainee will engage in three primary activities: 1. AI-Driven Protocol Development: Developing lightweight adapters using the Model Context Protocol (MCP) and REST APIs. The trainee will write asynchronous Python code to expose HPC and quantum capabilities securely to AI agents and interactive platforms (like Jupyter- JSC). 2. Hybrid Workflow Integration & Containerization: Packaging these service adapters into reproducible, containerised environments. The trainee will build end-to-end pipelines where an AI-driven interface successfully dispatches and retrieves jobs across Slurm/UNICORE for classical tasks and JUNIQ platforms for quantum tasks. 3. Benchmarking & Optimization: Systematically profiling the integrated workflows. The trainee will measure data serialization overhead, communication latency between AI agents and compute tiers, and validate the overall robustness of AI-orchestrated execution under realistic scientific scenarios. Through these activities, the trainee will directly contribute to lowering the operational barriers for computational research by enabling modern AI tools to seamlessly interact with Europe's premier supercomputing and quantum infrastructure.