Principal Scientist

Overview

About The Role
The organization seeks a Principal Scientist to lead the development of advanced computer vision and machine learning pipelines to translate high-resolution images of analytes into clinically validated diagnostic signals. This is a senior, hands-on industry role for a scientist who can translate medical images into clinically meaningful, validated signals and partner closely with translational researchers and Sonder's AI team to define analytical requirements, guide model development, and ensure imaging-derived outputs are translated into robust, scalable AI workflows.

The role involves designing and implementing computer vision algorithms for automated object detection, image segmentation, and feature extraction from microfluidic imaging data. The ideal candidate brings deep experience developing imaging analytics in commercial or clinical environments and is comfortable shaping capabilities, workflows, and standards that are still being built. This is a senior individual contributor role for a scientist who enjoys both shaping strategy and building production-grade imaging and AI workflows.

• Imaging strategy and scientific ownership
  Guide the scientific and analytical strategy for medical image analysis across the platform, serving as the internal subject-matter expert for imaging-derived signals.
• Image-derived biomarkers and ground-truth
  Serve as the primary subject-matter expert for medical imaging and image-derived biomarkers, informing scientific direction, technical tradeoffs, and long-term imaging strategy.
• Biomarker development
  Lead development of image-derived biomarkers, including radiomics and deep learning-based feature extraction from complex, high-resolution imagery such as pathology, satellite, or microscopic data.
• Annotation and ground-truth frameworks
  Define clinically grounded approaches to feature selection, annotation strategies, and ground-truth frameworks in collaboration with pathologists and clinical partners.

• Modeling and deployment readiness
  Provide scientific and clinical context to support model architecture decisions, training strategies, and interpretation of outputs. Ensure imaging pipelines are scalable, reproducible, and suitable for deployment as part of broader AI systems.

• Data integration
  Lead integration of imaging-derived features with molecular, clinical, and other data modalities as capabilities are established and expanded. Partner with scientific leadership to help define data requirements, analytical frameworks, and modeling strategies for a future multimodal diagnostic platform. Contribute to scalable, interpretable approaches for combining imaging with proteomic, transcriptomic, and other omics data as programs mature.

• Regulated development
  Translate imaging analytics into product-ready components, helping establish design controls, traceability, and risk-aware development practices as programs evolve. Experience with analytical verification and validation studies for AI-driven imaging workflows, ensuring the platform meets performance, robustness, and traceability standards required for IVD regulatory submissions. Support analytical validation planning and technical documentation for future clinical and regulatory requirements.

• Leadership
  Serve as a senior scientific partner to the AI team, aligning imaging analytics with machine learning workflows, timelines, and technical constraints. Provide technical guidance and mentorship to scientists and AI engineers as the organization grows, establishing strong standards for analytical rigor and reproducibility.

• Advanced degree (PhD strongly preferred) in biomedical engineering, computer science, medical physics, applied mathematics, or related quantitative field.
• 4+ years of relevant experience with medical imaging analytics, with demonstrated senior ownership in industry or commercial environments (healthtech, medtech, diagnostics, pharma, or equivalent).
• Deep expertise in medical imaging using computer vision; familiarity with pathology-relevant modalities is a strong plus.
• Workflows for interpretable machine learning
• Strong hands-on experience building and validating imaging ML models using modern tool chains (Python required; deep learning frameworks such as PyTorch or Tensor Flow).
• Experience working with medical imaging data and associated metadata; familiarity with standards such as DICOM and clinical imaging workflows is a plus.
• Demonstrated ability to collaborate effectively with clinical partners (for example, pathologists and, where applicable, life scientists), translating clinical needs into technical requirements and interpretable outputs.

• Experience with analytical verification and validation studies for machine learning/AI-driven imaging workflows, ensuring the platform meets the…