(1 year 5 months)

(4 years 9 months) [Acquired by Oura] Drove R&D efforts in finding highly reliable features in diverse physiological datasets using unsupervised learning techniques (e.g. contrastive learning) Developed time series CNN using Python / Tensorflow for event detection for human movement on force sensors Created over 50 metrics for biosensors using statistical signal processing, nonlinear dynamics, and physiological domain knowledge Built time series analysis products including anomaly detection and trend detection via online regression modeling.

Responsible for characterizing time series signal quality and denoising for a continuous glucose monitor

(1 year 11 months) Leveraging signal processing, machine learning, and deep learning techniques to develop mobile sensing detection algorithms. Projects include development of novel vehicle detection algorithms and pedestrian dead-reckoning. Developed novel algorithm to track phone's position in body coordinate system using biomechanical principles and sensor fusion. 2018 Mobile OS Employee of the Year

Led effort to develop machine learning seizure detection models based on clinical EEG data.

(1 year 4 months) End-to-end creation of advanced signal processing and machine learning algorithms related to physiological sensing of wearable devices, from data collection to implementation in C. Including: PPG-based Heart Rate Variability algorithm - robust to artifact. Adaptive & personalized walking / running step length algorithms. Includes novel GPS signal quality metrics. Novel algorithms for estimation of human power output during exercise. Patents and invention disclosures on various advanced physiological algorithms.

(1 year 8 months) Managed team of algorithm & firmware engineers for developing continuous HR, HRV, & activity detection deployed in a wearable sensor product Developed continuous HR estimation tracking algorithms based on advanced DSP techniques Intricately involved with hardware engineers in assessing and maximizing PPG signal quality

(3 years 1 month) Human injury assessment through experimental data analysis and computational modeling.

5 years 1 month (5 years 1 month) Designed and validated dynamic models using Matlab and ADAMS to test hypotheses related to human motion, including evaluating wheelchair propulsion seating and mechanics and landing control.

(4 years 5 months) Taught laboratory sections for the following courses: EXSC 205 - The Science of Human Performance EXSC 301 - Functional Anatomy EXSC 408 - Introduction to Biomechanics