Role overview

WHOOP is seeking a Staff Applied Machine Learning Scientist to develop and enhance production-ready algorithms that convert wearable sensor data into accurate, real-time physiological insights. You will combine deep learning, machine learning, and signal processing to deliver value to WHOOP members while optimizing for accuracy, latency, power, and scalability across wearable platforms.

Key responsibilities

Design advanced algorithms that integrate signal processing, physiological modeling, machine learning, and deep learning to analyze physiological time-series and multimodal sensor data with emphasis on accuracy, robustness, and broad generalization.
Own the full development and ongoing refinement of edge algorithms that reliably transform sensor data into accurate, real-time physiological insights for production deployment.
Analyze large-scale wearable sensor datasets to identify performance gaps, characterize challenging conditions, and implement data-driven improvements to algorithms.
Establish rigorous evaluation methodologies, validation frameworks, and performance metrics to measure algorithm performance throughout development and after deployment.
Optimize algorithms for embedded systems by balancing accuracy against constraints in power consumption, memory, latency, and computational resources across current and future wearable devices.
Define technical direction and establish best practices for modeling, experimentation, validation, and algorithm development across the organization’s sensing challenges.
Collaborate with Data Science, Firmware, Software, Hardware, Product, and domain expert teams to move algorithmic innovations from concept through production deployment and member-facing features.

Requirements

Master’s or doctoral degree in Electrical Engineering, Biomedical Engineering, Computer Science, Machine Learning, Applied Mathematics, or a closely related quantitative discipline.
Seven or more years of professional experience developing and deploying machine learning, deep learning, and signal processing algorithms for complex real-world applications.
Deep technical expertise in contemporary machine learning and deep learning methods, with particular strength in time-series and multimodal sensor data.
Strong background in digital and statistical signal processing with demonstrated ability to combine classical signal processing with modern learning-based approaches.
Strong proficiency in Python for algorithm development, experimentation, and large-scale data analysis; C/C++ and embedded algorithm development experience is highly desirable.
Proven ability to develop robust models using large, noisy, real-world datasets that generalize well across diverse conditions.
Experience designing rigorous experiments, defining metrics, performing error analysis, and translating insights into algorithm enhancements and experience advancing algorithms through complete development cycles from research and prototyping through validation, deployment, monitoring, and continuous improvement.

What makes this role worth considering

This position is based in Boston, MA, where you will work directly with cross-functional teams of scientists, engineers, physiologists, and hardware experts on challenging problems at the intersection of wearable sensing, physiological modeling, and applied AI. Your work will directly influence WHOOP’s sensing capabilities and shape the company’s ability to deliver accurate, personalized, and actionable health and performance insights to members.