I work at the intersection of human movement science, physiology, and wearable sensing. I design studies, build practical measurement workflows, and help teams make defensible claims from real-world data. ML is not my main selling point—rigorous human-centered measurement and clear translation are.
IMU-based movement analysis and multi-sensor setups (EMG/ECG/PPG-style streams), with practical attention to placement, artifacts, drift, and real-world constraints—including emerging form factors like smart glasses and earables.
Designing studies teams can execute: protocols, endpoints, feasibility, adherence, and data-quality checks. Experience across observational and intervention contexts, translating questions into testable outcomes.
Domain depth in gait and functional tasks, neuromuscular and physiological interpretation, and what movement/physiology signals can (and cannot) support in real humans.
Working with large wearable datasets (10k+ participants) to derive interpretable activity-pattern metrics (bouts, fragmentation, transitions, intensity distribution) and relate them to cardiometabolic outcomes.
“Lab-to-life” benchmarking plans: gold-standard comparisons, synchronization, artifact handling, error reporting, and transparent limitations—so performance claims are defensible and useful for product decisions.
A few compact examples (Problem → Approach → Output). I can share deeper case-study PDFs or references on request.
Problem: Large accelerometry datasets are messy, and “steps/day” misses meaningful behavior.
Approach: Built reproducible Python workflows to quantify pattern metrics (bouts, fragmentation, transitions, intensity distribution),
with data-quality checks and reporting built for decisions—not just plots.
Output: Analysis-ready measures and clear summaries to support links with cardiometabolic outcomes.
Problem: Lab snapshots often miss day-to-day movement strategies during rehabilitation.
Approach: Designed a wearable workflow for functional tasks (multi-pace walking and beyond), emphasizing signal quality,
event logic, and outputs clinicians and engineers can both interpret.
Output: Reusable pipeline components and figures suitable for stakeholder reporting and publication-grade materials.
Problem: Novel form factors need credible benchmarking against reference measures in realistic conditions.
Approach: Designed a synchronized data-collection protocol across daily activities with practical constraints in mind
(comfort, adherence, time alignment, artifacts), and a clear analysis/reporting plan.
Output: Validation-ready protocol and roadmap for translating signals into defensible performance claims.
Problem: Using wearable-derived outcomes responsibly requires clear measurement assumptions and QA.
Approach: Defined outcomes and checks up front, aligned analysis to what the sensor can support, and emphasized transparent limitations.
Output: A practical framework for using wearable kinematics as credible outcomes in applied human research settings.
Problem: Can accessible devices produce research-grade measurements with standardized protocols?
Approach: Reliability testing with standardized positioning, clear reporting of error, and transparent constraints.
Output: A practical measurement approach designed to survive real-world clinical variability.
I’m open to opportunities across Applied Research, R&D, Measurement/Validation, and technical project coordination—especially in wearable sensors and assistive systems (including exoskeleton-adjacent work).
Get in Touch