Responsibilities

Design and implement pre-processing pipelines for signal and time series data.

Build algorithms/models for heart rate, respiration rate, movement detection, and

sleep stage estimation.

Develop robust algorithms for real-world noisy environments and validate them against

reference datasets

Conduct feature engineering for sleep-related ML models (time/frequency/wavelet

features).

Collaborate with firmware and cloud teams to integrate algorithms into on-device and

cloud pipelines.

Contribute to internal signal quality, scoring, and annotation tools.

You Have

45 years of experience in signal processing/data science/biomedical sensing.

Solid foundation in digital signal processing (filters, FFT, IIR/FIR, wavelets).

Experience with image models(CNN) and sequence/time series models

Proficiency in Python (NumPy, SciPy, mne, neurokit2, pandas, matplotlib,

sckit-learn)

Experience with sleep staging, physiological data analysis, or similar time-series

modeling.

Comfort with real-world data: noise, motion artifacts, gaps, resampling, and validation.

(Bonus) Experience with embedded signal processing, PyTorch/TF for physiological

ML, or on-device inference.

Nice to Have

Prior work with wearables, health tech, or medical devices (Ultrahuman, Dozee,

Fourth Frontier, Muse, Fitbit).

Familiarity with BCG datasets, wavelet-based HR/RR detection, or open-source

sleep datasets (PhysioNet, ISRUC, Sleep-EDF).

Understanding of biosignal labeling, ML evaluation, deployment and monitoring

EPM Consultant