Senior Data Scientist, Imaging Foundation Models

Merck Merck · Pharma · MA

Senior Data Scientist role at Merck focused on developing and training imaging foundation models for histopathology, with a focus on immune-relevant tissue and cellular patterns. The role involves building segmentation-driven and cell-centric pipelines, designing interpretable modeling workflows, and applying zero-shot and weakly supervised learning methods. Collaboration with multidisciplinary teams and contribution to platform strategies are key.

What you'd actually do

  1. Develop and train imaging foundation models for histopathology, with a focus on immune‑relevant tissue and cellular patterns.
  2. Build segmentation driven and cell‑centric pipelines (e.g., immune cell typing, spatial organization, microenvironment analysis).
  3. Design interpretable modeling workflows that link image derived features to biological hypotheses and clinical endpoints.
  4. Apply zero‑shot and weakly supervised learning methods to extract immune signals from H&E and IHC data.
  5. Collaborate closely with immunologists, pathologists, and translational teams to ensure biological relevance and interpretability.

Skills

Required

  • PhD in Computer Science, Engineering, Data Science, AI/ML, Bioinformatics, Computational Biology, Genetics & Genomics, Mathematics, Statistics, Physics, Pharmaceutical Science, or related STEM field with 0+ years postdoctoral experience or Master’s degree with 4+ years of industry experience.
  • Strong expertise in computer vision and medical image analysis and/or multimodal data.
  • Experience building and/or applying models for segmentation, detection, and representation learning, ideally in histopathology.
  • Familiarity with modern deep learning architectures (e.g., transformers, vision foundation models).
  • Proficiency in Python and deep learning frameworks such as PyTorch.
  • Strong interest in immunology and tissue‑based biomarker discovery.
  • Ability to communicate complex technical ideas to multidisciplinary scientific partners.

Nice to have

  • Familiarity with clinical or translational imaging biomarker programs.
  • Publications in venues such as MICCAI, CVPR, NeurIPS, ISBI, or related journals.

What the JD emphasized

  • Strong expertise in computer vision and medical image analysis and/or multimodal data.
  • Experience building and/or applying models for segmentation, detection, and representation learning, ideally in histopathology.
  • Familiarity with modern deep learning architectures (e.g., transformers, vision foundation models).

Other signals

  • foundation models
  • biomedical imaging
  • computer vision
  • interpretable pipelines