• Built algorithms used to calculate features for various machine learning classifiers and models. Algorithms include methods to calculate normal r-r intervals in low sample-rate PPG signals, detect Mayer waves on irregularly sampled heart rate variability estimates, and composite bioimpedance pulse waveforms for aggregate wavelet analysis.• Wrote and optimized library of fixed-point DSP operators used as a part of a real-time algorithm for predicting heart rate in wearable devices running on ARM microprocessor.• Led internal testing integral to making product decisions. Responsibilities included aligning functional groups on success criteria, developing testing protocols, then analyzing and presenting results.• Implemented and maintained data pipeline to combine and time-align disparate sensor data streams into singular data formats stored on AWS, easily accessible throughout the organization.