Role Overview
Join a forward-thinking team focused on transforming cancer diagnostics through artificial intelligence. In this role, you'll design and implement deep learning systems for analyzing whole-slide images, bridging the gap between research innovation and real-world clinical application. Your work will directly contribute to building reliable, scalable tools that support pathologists and improve patient care.
Key Responsibilities
- Develop and deploy deep learning models for whole-slide image analysis, with applications in object detection, segmentation, and slide-level classification—particularly in oncology workflows such as prostate biopsy assessment.
- Lead efforts to improve model accuracy, calibration, and interpretability by building visual explanations, region-of-interest summaries, and tools that support clinical validation and quality assurance.
- Create robust methods to handle real-world challenges in digital pathology, including variations in staining, scanner types, and data drift over time.
- Own the full AI lifecycle—from research and prototyping to production deployment and ongoing performance monitoring.
- Design and run experiments that inform model development, including ablation studies, benchmarking against state-of-the-art methods, and detailed error analysis.
- Deliver production-ready implementations with attention to model serving, performance tracking, and feedback loops for continuous improvement.
- Partner with software engineers to build scalable, containerized systems and APIs that integrate smoothly into existing clinical workflows.
- Collaborate with pathologists to define clinically relevant evaluation metrics, concordance standards, and adjudication processes.
- Support regulatory and quality requirements for medical devices, including documentation, traceability, and validation aligned with clinical lab standards.
- Mentor team members, lead code reviews, and promote engineering best practices across the AI function.
- Help shape the long-term technology roadmap by integrating emerging research and advancing the platform’s capabilities.
- Communicate technical findings clearly to cross-functional teams, including product, regulatory, and clinical partners.
- Evaluate trade-offs across different AI approaches with a pragmatic, context-sensitive mindset.
- Stay current with advancements in computer vision and computational pathology to ensure the platform remains at the leading edge.
Required Qualifications
- At least 6 years of hands-on experience building deep learning models for computer vision, with proven ownership of at least one system deployed in production.
- Strong command of Python and PyTorch, with deep knowledge of modern machine learning and computer vision techniques.
- Commitment to software engineering excellence—writing clean, testable, maintainable code, practicing code reviews, and following CI/CD and reproducibility standards.
- Experience working in Linux environments and with containerized development workflows, including model deployment and monitoring in production settings.
- Ability to design evaluation frameworks that reflect real-world clinical performance and usability.
- Practical experience using cloud platforms such as AWS, GCP, or Azure for scalable AI deployments.
Preferred Qualifications
- Degree in computer science, engineering, or a related field—Master’s or PhD preferred.
- Background in medical imaging, especially digital pathology and whole-slide image analysis.
- Familiarity with tools commonly used in computational pathology, such as QuPath, CellProfiler, or OpenSlide.
- History of publishing research or contributing to applied AI products with real-world impact.
- Experience with MLOps platforms like MLflow, and practices such as data versioning, model drift detection, and performance regression tracking.
- Knowledge of domain generalization, stain normalization, multi-scanner compatibility, and quality control in histopathology imaging.
- Understanding of regulated environments, particularly in medical devices or IVDs, with attention to risk management, documentation, and traceability.
Technical Environment
PyTorch, OpenCV, Scikit-learn, Scikit-Image, Pandas, TensorBoard, Git, Docker, Bash
