Senior Data Scientist

Many structural heart patients suffer from heart failure with limited options. Our Implantable Heart Failure Management (IHFM) team part of the AI Product and Platforms organization is at the forefront of addressing these unmet patient needs through pioneering technology that enables early targeted therapeutic intervention. Our innovative solutions are not just transforming patient care but also creating a unique and exciting environment for our team members. It is our driving force to help patients live longer and healthier lives. Join us and be part of our inspiring journey.

At Edwards Lifesciences the Implantable Heart Failure Management (IHFM) AI Product and Platforms organization designs and builds the software and data products that clinicians and patients depend on. As a Senior Data Scientist you own modeling for a clinical or product domain drive problem framing with product and clinical partners own offline-validation rigor and author model documentation that meets FDA submission expectations for regulated products.

Based in Irvine CA youll join a high-impact medtech innovation hub in the heart of Orange County collaborating in person with cross-functional teams to shape patient-focused technology.

• Domain modeling. Own modeling for a clinical or product domain for example arrhythmia classification cardiac imaging segmentation clinical prediction or patient outcome forecasting from framing through offline validation.

• Architecture selection. Select and adapt architectures for the modality (vision transformers or encoder-decoder networks for imaging transformers or temporal models for signals) and apply self-supervised and representation learning under limited labels.

• Statistical depth. Apply survival analysis Bayesian modeling and causal inference where the clinical question requires (statsmodels lifelines or scikit-survival PyMC).

• Evaluation & subgroup. Design offline evaluation calibration and clinical performance validation strategies including subgroup and bias analysis across sex age and device cohort with Medical Affairs and Clinical Science.

• Interpretability & documentation. Author submission-grade model documentation including calibration uncertainty and interpretability evidence (SHAP Captum) and improve the teams modeling and validation practice.

• Research adoption. Evaluate and adapt current AI/ML for health research to IHFM problems and mentor less experienced data scientists.

• Handoff. Coordinate the handoff of offline-validated custom models to AI/ML Engineers (Applied) and partner with Regulatory Affairs on submission strategy.

• Algorithm development analysis & reporting. Create test and improve complex algorithms NLP and machine learning models; analyze results; develop insights; and produce reports and dashboards to communicate performance and findings to stakeholders.

• Data preparation & quality. Process cleanse label and verify structured and unstructured data used for analysis; collaborate with internal teams and external partners on data standards metrics analytics and reporting.

• Tools integration & design controls. Identify and integrate data sources software and analytics tools (e.g. Python SQL SAS Power BI Tableau Prep); support development of design control documentation including algorithm requirements and risk documentation.

• Bachelors in related field (e.g. Computer Science Engineering Biostatistics or Scientific) plus 4 years -or- Masters plus 3 years of previous experience including industry or industry/ education

• This is an onsite role based in Irvine CA. Relocation assistance is not provided and candidates must reside within a 50-mile radius of Irvine to be considered.

• Depth in at least one data modality relevant to IHFM for example time-series or medical imaging (echocardiography cardiac magnetic resonance imaging or computed tomography).

• Command of the relevant architecture families and of transfer and self-supervised learning.

• Strong Python and SQL and depth in modeling and experiment tracking such as PyTorch TensorFlow and MLflow.

• Command of evaluation calibration and clinical performance methodology including subgroup analysis and the ability to author documentation suitable for regulatory submission.

• A track record of owning modeling for a domain and communicating results to clinical product and regulatory partners.

• Experience with clinical validation retrospective clinical data or regulated medical software (SaMD).

• Depth in time-series or medical imaging modeling including segmentation and registration for example MONAI pydicom or SimpleITK.

• Bayesian modeling causal inference (DoWhy EconML) or survival analysis (lifelines scikit-survival).

• Generative approaches for augmentation or synthetic data (autoencoders GANs or diffusion models).

• Exposure to multimodal modeling combining signals imaging labs and notes into patient-state models.

• Familiarity with R JAX or hyperparameter optimization (Optuna Ray Tune).

• Contributions to internal or external research (publications patents or venues such as MICCAI or ML4H).

Aligning our overall business objectives with performance we offer competitive salaries performance-based incentives and a wide variety of benefits programs to address the diverse individual needs of our employees and their families.

For California (CA) the base pay range for this position is $108000 to $153000 (highly experienced).

The pay for the successful candidate will depend on various factors (e.g. qualifications education prior experience). Applications will be accepted while this position is posted on our Careers website.

Edwards is an Equal Opportunity/Affirmative Action employer including protected Veterans and individuals with disabilities.