Staff Applied Machine Learning Scientist (Health)
A Staff-level applied ML scientist on WHOOP's Health team who designs, validates, and sets methodological standards for machine learning and statistical methods behind health features, working with longitudinal wearable physiological data (PPG, HRV, sleep, activity). On-site in Boston, partnering with ML engineers, product, and clinical teams to take models toward production.
- Compensation
- $170K–$230K / yr
- Location
- Boston, MA
- Work setup
- Onsite
- Seniority
- Staff
- Employment
- Not specified
- Experience
- 7+ years
- Category
- Ml / AI
- Posted
Skills
Must have
- Master's or PhD (or equivalent professional experience) in Biomedical Engineering, Electrical Engineering, Computer Science, Applied Mathematics, Biostatistics, Bioinformatics, Physiology, or a relate
- 7+ years of professional experience developing algorithms, applied ML models, or statistically rigorous analyses in health, wearables, medical devices, digital health, or a closely related domain
- Demonstrated Staff-level technical leadership: setting methodological standards others reuse, influencing multi-team outcomes, mentoring senior ICs, and making prioritization calls under uncertainty
- Strong foundation in statistics (Bayesian and frequentist) and machine learning for time-series and longitudinal data
- Experience with wearable or physiological signals (PPG, IMU, HR/HRV, sleep, activity, temperature, or similar) or a closely related adjacent time-series domain
- Hands-on experience designing studies or evaluation protocols end-to-end: data curation, metric choice, validation splits, subgroup and failure analysis, and documentation of assumptions and limitatio
- Strong Python skills with reproducible scientific/ML engineering hygiene (NumPy/SciPy/pandas, scikit-learn, PyTorch/TensorFlow)
- Clear, high-signal communicator able to move between deeply technical method discussions and cross-functional conversations with Product, clinical partners, and leadership
- Willing to relocate if necessary to work out of the Boston, MA office
Full description
At WHOOP, we’re on a mission to unlock and inspire performance for life. WHOOP empowers members to perform at a higher level through a deeper understanding of their bodies and daily lives. Our wearable technology collects rich physiological data, providing members with actionable insights into their recovery, training, and sleep.
The Health team develops the algorithms and features that expand WHOOP's health sensing capabilities. The work spans women's health, software as a medical device, wellness and longevity, member insights, and emerging health signals. We combine continuous physiological data with clinical research and domain expertise to ship features that are scientifically grounded and useful to millions of members.
As a Staff Applied Machine Learning Scientist on the Health team, you will turn ambiguous clinical and product questions into defensible studies, models, and evaluation plans on WHOOP's longitudinal wearable and member data (physiological signals, behavioral logs, labs and integrations, and related streams). You will set the methodological bar for a Health domain. You will partner with Machine Learning Engineers on the path to production, and with Product, clinical, and Digital Health partners on what the data can and cannot support.
Success in this role requires applied algorithm and statistical craft, the judgment to operate across Health problem types, and Staff-level leadership: launching domain work independently, raising the quality and velocity of the people around you, and making clear prioritization calls under uncertainty.
Responsibilities
- Lead the design, development, and validation of machine learning and statistical methods for Health features, using wearable physiological and behavioral data and related member sources.
- Frame ambiguous health and product questions into well-scoped modeling or study problems, with clear performance criteria, evaluation plans, and honest treatment of uncertainty and failure modes.
- Set the methodological and quality standard for a Health domain: reusable evaluation practices, review expectations, and documentation that others can inherit.
- Develop and refine approaches for time-series, longitudinal, and multimodal data, including data curation, labeling strategy, feature and representation choices, and offline evaluation that holds up under real-world noise, missingness, and population shift.
- Partner with Machine Learning Engineers to translate validated methods into production-ready systems, making modeling choices that account for scalability, reliability, latency, cost, monitoring, and (where relevant) controlled-release constraints.
- Serve as a technical counterpart to Product and Digital Health (clinical science, regulatory, quality): pressure-test performance requirements, study design, and evidence quality, and push back when a claim is not supportable.