Research Assistant DECIPHER (m/f/d)

Development and implementation of interfaces between clinical data sources, medical knowledge bases and local LLM/Agent systems Build powerful data pipelines for loading, structuring, validation and reprocessing of interoperable health data Development of the technical architecture of LLM agents Integration of structured and unstructured clinical information into agentic workflows for personalized cancer prevention and screening recommendations Development of APIs, data models and software components for secure communication between clinical systems, AI modules and user interfaces Conception and implementation of local on-premise deployments of LLM-based components taking into account data protection, IT security, traceability and clinical usability Connection and further development of knowledge bases, retrieval-agugmented generation components and guidelines modules for medically comprehensible recommendations Development of user interfaces and demonstrators for clinical users, study sequences, shadow evaluations and feedback iterations Systematic evaluation of the developed interfaces and AI components with regard to technical robustness, efficiency, medical plausibility, provenance, guidance and applicability in clinical everyday life Close cooperation with clinical and technical project partners to translate clinical requirements into robust software solutions Documentation of developed software, interfaces and data flows as well as participation in scientific publications, conference contributions and project reports

Graduate degree in Computer Science, Artificial Intelligence, Engineering, Data Science or a related subject Very good programming knowledge in Python as well as experience in the development, structuring and maintenance of complex software projects Knowledge of machine learning, data science and modern methods of artificial intelligence Practical experience with Large Language Models, agentic AI systems, Retrieval-Augmented Generation, Prompting, Model Evaluation or comparable applications is explicitly desired Experience with the development of data pipelines, interfaces and modular software architectures, ideally for heterogeneous or sensitive data Interest in local, safe and resource-efficient deployment of AI models, including inference, hardware requirements, latency and scalability issues Ability to document, systematically test and scientifically evaluate technical solutions Extensive interest in in interdisciplinary research at the interface of AI, software development, medicine, clinical implementation and trustworthy AI Very good knowledge of English in word and writing as well as knowledge of German, especially for cooperation with clinical users

in the Else Kröner Fresenius Centre for Digital Health Project Deciphering Metastasis with Multimodal Artificial Intelligence Foundation Models At the next possible time, the body must first be fixed for 24 months in full or part-time employment, with the option of extension and longer-term cooperation. Remuneration is carried out in accordance with the classification rules of the public service collective agreement (TV-L) and is possible in the case of personal conditions in the E13 remuneration group. Remuneration: according to TV-L in the remuneration group E13 plus Annual special payment, in the event of personal conditions Holiday: 30 days Recreational holiday Continuing training: Extensive internal and external continuing and continuing training opportunities to continuously strengthen your skills Corporate culture: We focus on transparency, team spirit and common successes. Celebrate with us at joint events or at sporting events Mobility: Grant to the job ticket to make your work path environmentally friendly and convenient Health: We promote your well-being with modern sports and exercise offerings in our health center, mental health programs and comprehensive occupational health management Benefits: Benefit from staff rebates in our clinic pharmacy, exclusive corporate benefits and other shopping portals Work-Life-Balance: Our family office supports you with questions about Kita courses, holiday programs and care of relatives. We also offer flexible working models