Postdoctoral Researcher in Confidential High Performance Computing for AI in Cancer Care University of Basel in Switzerland

The University of Basel invites applications for a Postdoctoral Researcher in Confidential High Performance Computing for AI in Cancer Care, based in the High Performance Computing (HPC) research group led by Prof. Florina M. Ciorba in the Department of Mathematics and Computer Science. This position is funded by the SNF Bridge Discovery project 'Family Gene Toolkit (FGT)', a digital platform supporting families affected by hereditary cancer syndromes such as HBOC and Lynch Syndrome. The next-generation FGT v3.0 aims to integrate large language models, multilingual AI, and confidential HPC to support patients and clinicians in real-world healthcare environments.

The successful candidate will lead research on trustworthy and aligned large language models (LLMs) operating in confidential HPC environments, including training and adapting 70B-scale multilingual and medical language models on secure HPC infrastructure. Key research areas include preference optimization and LLM alignment (RLHF, PPO, DPO, reward modeling), agentic and tool-augmented AI systems (reasoning, RAG, in-context learning, chain-of-thought reasoning, agentic workflows), and trustworthy AI evaluation and confidential deployment (evaluation pipelines, robustness, bias/hallucination mitigation, multilingual evaluation protocols).

Responsibilities include conducting high-quality research aligned with project objectives, publishing in top venues (NeurIPS, ICML, ICLR, ACL/EMNLP, MLSys, SC, HPDC, EuroSys, ASPLOS), presenting at international conferences, contributing to open-source software and reproducible research artifacts, collaborating with the HPC group and partner institutions, contributing to teaching (max. one class per semester), and co-supervising Bachelor's/Master's students.

Applicants must hold a PhD in Computer Science, AI, or Machine Learning, with a strong publication record in AI, ML systems, or related areas. Essential skills include programming in Python, C/C++, experience with PyTorch, TensorFlow, JAX, or similar ML frameworks, experience with large-scale training or inference of LLMs, interest in LLM alignment, reinforcement learning, or generative AI systems, and fluency in English. Highly desirable are experience with distributed training or ML systems, knowledge of privacy-enhancing technologies and parallel programming, and experience with multilingual AI or agentic systems. The group welcomes candidates from diverse backgrounds and encourages applications even if not all requirements are met.

The position offers competitive 100% funding per SNSF guidelines (~CHF 90,000/year), access to modern GPU clusters and confidential-computing infrastructure, collaboration with leading researchers in AI, HPC systems, and digital health, support for international conference travel and networking, and a stimulating research environment at a leading European university.

To apply, submit a single PDF file via the application portal containing your CV with publication list, degree transcripts, theses, links to code repositories or software projects, at least one relevant publication with justification, a motivation statement, and contact details for 1–2 referees. Applications by e-mail will not be considered. Only short-listed candidates will be invited for interview. The application deadline is April 24, 2026, with rolling review until filled.

For questions, contact Prof. Florina M. Ciorba at [email protected] with the subject 'BRIDGE FGT – Postdoc position'.