Your position
You will design, develop, and integrate AI-based and privacy-preserving features into FGT v3.0 within a confidential HPC environment, including :
- Train and fine-tune large medical LLMs (70B-scale) on confidential HPC systems.
- Prepare medical datasets (Clinical Guidelines, NCCN, PubMed, MIMIC) for secure LLM training.
- Develop high-performance pipelines for TEE-based model training and inference.
- Benchmark models using standard medical and multilingual evaluation suites.
- Build prototype chatbot and retrieval systems for clinical partners.
- Integrate medical LLMs with multilingual models (Apertus).
- Contribute to robustness and confidential-AI evaluation workflows.
Your tasks also include
- Conduct research aligned with project objectives
- Write and submit high-quality papers to leading conferences and journals in HPC and AI, for healthcare
- Present results at seminars, workshops, and international conferences
- Interact actively, fruitfully, and respectfully within the PI, HPC group, and across partner institutions
- Contribute to teaching as assistant (one class per semester)
- Contribute to supervision as assistant (one student or more per semester)
Your profile
Requirements for the position
- Master's degree in Computer Science / Engineering or a closely related field at the start date (March 2026 at the latest)
- Excellent academic record
- Programming skills in C, C++, Java, or Python
- Experience with parallel programming, Linux, machine-learning frameworks, and privacy-enhancing technologies
- Fluency in English (spoken and written)
- Clear communication, problem-solving ability, and collaborative mindset
- Genuine curiosity for Computer Science and motivation to conduct rigorous research with impact in Computer Science and Healthcare