AI/ML Engineer, Clinical AI

About IDM

IDM gGmbH is a nonprofit health technology company developing AI tools tailored for clinical environments. Their technology is operational at the University Medical Center Hamburg-Eppendorf (UKE), one of Europe's leading university hospitals, as well as at other hospital networks across Germany. Their main products include ORPHEUS, a Clinical-NLP platform for medical documentation, and ARGO, an integrated clinical workplace system. IDM bridges cutting-edge machine learning research and practical clinical applications, delivering systems actively used by medical professionals.

Role Overview

As a Senior AI/ML Engineer, you will manage the complete machine learning lifecycle for clinical AI models—from research and experimentation to deployment and monitoring in production. You will gain direct access to authentic clinical data at UKE Hamburg, a unique opportunity in the ML domain. This position combines hands-on engineering with research leadership, allowing you to shape the research agenda, contribute to clinical studies, publish in top venues, build a research team, and ensure your models impact patient care.

Key Responsibilities

• Design, train, and optimize machine learning models for clinical NLP applications involving medical documentation, clinical decision support, and medical text analysis using real clinical datasets.
• Develop and maintain end-to-end ML pipelines covering data preparation, training, evaluation, deployment, and production monitoring within hospital environments.
• Lead clinical AI research projects in collaboration with university hospitals, contribute to publications, and assist in securing external research funding from sources such as government bodies and EU programs.
• Mentor students and supervise scientific assistants, fostering the growth of IDM's AI research efforts.
• Advance MLOps capabilities including model versioning, automated training pipelines, experiment tracking (MLflow, W&B), and robust deployment workflows.
• Optimize model architectures for clinical deployment focusing on latency, scalability, and resource efficiency in real-time settings.
• Develop strategies for handling data quality, anonymization, and harmonization across heterogeneous clinical data types including text, imaging, time series, and genomics.

Required Qualifications and Skills

• Minimum of three years hands-on experience in AI/ML with a strong emphasis on natural language processing.
• Proficient in Python and at least one major ML framework, preferably PyTorch; TensorFlow experience is acceptable.
• Familiarity with the HuggingFace ecosystem, including Transformers, Datasets, and Tokenizers.
• Strong understanding of contemporary NLP architectures such as Transformers, large language models, and encoder-decoder models.
• Experience with MLOps tools and practices, including MLflow, Docker, and CI/CD pipelines tailored for ML.
• Solid software engineering skills including version control with Git, testing, code reviews, and documentation.
• Demonstrated ability to independently manage projects from concept through deployment.

Preferred Additional Qualifications

• Publications in top ML or NLP conferences (e.g., NeurIPS, ICML, ACL, EMNLP, NAACL) or medical AI journals.
• Experience working with clinical or medical data and healthcare AI applications.
• German language skills beneficial but not mandatory for clinical collaboration.
• Advanced academic degree such as a Master's or PhD in Computer Science, Machine Learning, Computational Linguistics, or related disciplines.

Technical Environment

• Core ML technologies: Python, PyTorch, HuggingFace Transformers, scikit-learn, spaCy.
• MLOps tools: MLflow, Weights & Biases, Docker, Kubernetes, GitLab CI.
• Infrastructure components: PostgreSQL, S3 storage, Helm, Grafana, Prometheus.
• Computing resources: CUDA-enabled GPU clusters hosted on STACKIT Cloud.

What IDM Offers

• Exclusive access to genuine clinical datasets—a rarity in the machine learning community.
• Freedom to influence the research agenda and publish your findings.
• A mission-driven environment focused on improving healthcare without commercial advertising interests.
• A close-knit senior team where your contributions have direct impact.
• A hybrid work model with an office located in Hamburg-Eppendorf directly adjacent to UKE.
• Flexible working hours and a transparent, collaborative workplace culture.
• Funding for professional development and attendance at conferences.