Senior Machine Learning Engineer

Job Summary

The Senior Engineer, R&D is responsible for developing, improving, and delivering machine learning models for DeepHealth clinical AI products. This hands-on role spans data, experimentation, model development, evaluation, and production delivery, working with machine learning peers, software engineers, clinicians, and product partners to investigate problems, make technical decisions, and deliver measurable improvements in model quality, robustness, and operational performance.

Essential Duties and Responsibilities

Improve existing production models through systematic error analysis, better data, targeted experiments, and changes to model architecture and training.

Develop models for new products, taking problems from initial formulation and feasibility experiments through training, validation, and production integration.

Partner with clinicians and product colleagues to define meaningful evaluation criteria, including sensitivity, specificity, and the clinical consequences of different error types.

Evaluate robustness across patient populations, clinical sites, imaging equipment, and acquisition conditions; identify performance gaps and build evidence that improvements generalize.

Improve data curation and annotation workflows, including coverage gaps, label quality, and prevention of data leakage.

Build reproducible training and evaluation pipelines with traceable datasets, experiments, and model versions.

Partner with software engineers to optimize inference speed, resource use, and operational reliability, and investigate model issues that emerge in production.

Review relevant research, test promising approaches, and make evidence-based decisions about what to adopt.

Contribute to validation and technical documentation with quality and regulatory colleagues.

Review code and experiments, mentor colleagues, and communicate findings and trade-offs clearly.

Bachelor's degree in computer science, engineering, mathematics, or a related field, or equivalent practical experience (required).

5+ years of hands-on experience developing and delivering machine learning models, with evidence of independently taking complex work from an initial problem to a working solution (required).

Strong foundations in deep learning and computer vision, including practical experience with image classification, detection, or segmentation (required).

Strong Python skills and experience with a modern deep learning framework such as PyTorch (required).

Track record of deploying models into products and measuring performance beyond development datasets (required).

Rigor in experimental design and evaluation, including appropriate baselines, uncertainty, failure-mode analysis, and distinguishing meaningful gains from noise (required).

Strong software engineering practices, including maintainable code, testing, version control, and reproducibility (required).

Sound judgement on trade-offs between model quality, complexity, inference cost, and delivery time (required).

Ability to work autonomously and collaborate effectively across disciplines, with clear written communication (required).

Preferred: Medical imaging experience, or other applications involving variable image quality and limited or noisy labels.

Preferred: Developing and validating models for regulated products.