Senior/Principal Scientist - Translational Statistics - Predictive Modeling

Servier

Gif-sur-Yvette, France

Position Overview:

As a key member of the Translational Statistics team, you will be responsible for generating quantitative evidence to support development strategies across the translational continuum. Operating at the interface of Translational Medicine, Clinical Development, and Biometrics, you will ensure that biomarker and translational research data are converted into actionable conclusions to inform strategic development choices, from preclinical research to early clinical studies.

Main Responsibilities:

Translate biomarker and translational insights into decision-relevant evidence supporting development strategy.
Integrate preclinical, RWD, and early clinical evidence to inform PoC positioning and Go/No-Go considerations.
Quantify the credibility of biomarker-based enrichment strategies in data-limited PoC settings.
Develop predictive ML models to anticipate biomarker-defined treatment benefit or toxicity with limited biomedical datasets.
Develop ML approaches to infer actionable biomarkers or proxy markers that facilitate identification and recruitment of targeted patients in clinical trials.
Oversee and critically review models developed by external AI partners, ensuring performance robustness, and alignment with intended context of use.

Required Skills and Experience:

Educational & Professional Background: PhD in Biostatistics, Data Science, Machine Learning or related quantitative discipline. MSc candidates with relevant experience could also be considered. 5+ years of experience applying statistical or machine learning methods to biomedical or pharmaceutical datasets.
Technical Expertise (Hard Skills):
- Solid understanding of early clinical development processes and decision frameworks.
- Statistical reasoning in small sample and high-uncertainty settings.
- Experience in ML for predictive modeling in limited data contexts (data augmentation, transfer learning, generative models, foundation models).
- Experience with model interpretability and translation of ML outputs into development decisions.
- Familiarity with integration of multi-modal biomedical data (clinical, molecular, digital…).
Language: Scientific level English.