Summary
✨ AI‑Generated
A growing healthcare technology company is seeking a full-stack engineer to develop data-driven applications and algorithms for connected medical solutions. You will work across backend systems, analytics pipelines, and user applications while transforming complex sensor data into meaningful insights.
Highlights
Opportunity to build end-to-end health technology solutions, combining software engineering, data science, and advanced analytics. The role offers high ownership and the chance to create impactful features from raw data processing to user-facing applications.
Description
About AeroCardiaAeroCardia is a Montreal-based medical device startup developing a portable, multi-sensor cardiopulmonary monitoring system.
Our mouthpiece form factor integrates spirometry, PPG, SpO₂, CO₂, and inertial measurement into a single wearable device with a companion mobile app and clinical analytics subscription.
We're pre-launch for a direct-to-consumer wellness offering (Jan–Mar 2027) and pursuing parallel regulatory pathways with Health Canada (Class II ITA) and the FDA (510(k)) for clinical RPM applications.
Why this matters: VO₂max estimation is the cornerstone of our launch and customer value proposition.
You'll own the algorithms that convert raw sensor fusion data into validated, personalized cardiorespiratory insights.
The RoleYou are a full-stack data scientist / backend engineer who will own both algorithm development and the app backend that delivers VO₂max insights to users.
This is a rare opportunity to ship an end-to-end feature—from raw sensor fusion to user-facing mobile app—in a high-stakes medtech environment.
Algorithm & Data Science (50% of time)VO₂max model ownership: Refine and validate existing VO₂max models using HR data, spirometry, and inertial motion patterns; reduce systematic error (current SEE ~1.3–1.5 ml/kg/min) and improve robustness across user populationsSignal processing & feature engineering: Extract and optimize features from device sensor streams (DHF, PPG windowing, DFA-α1, VT detection via IMU-gait fusion)Validation & experimentation: Collaborate with our physiology team (Concordia testing lab) to design and interpret validation experiments; compare device output against reference standardsModel training pipeline: Build reproducible training/evaluation workflows in Python (scikit-learn, PyTorch, or TensorFlow); manage datasets, cross-validation, hyperparameter tuningBackend & App Integration (40% of time)Model deployment: Implement inference pipelines; package VO₂max models for iOS/Android (TensorFlow Lite, CoreML, ONNX, or similar)Backend API design: Build REST/GraphQL endpoints that serve VO₂max predictions to the app; handle device-to-cloud data sync, user authentication, and data persistenceReal-time processing: Design data pipelines that consume raw sensor telemetry from devices and compute VO₂max estimates in near-real-timeScalability & performance: Optimize for latency, cost, and reliability as we scale from beta (5/day) to 20+/day during UCI Worlds to thousands of active users post-launchCollaboration & Documentation (10% of time)Cross-functional work: Partner with iOS/Android developers (currently Matthew Pompeo), hardware/firmware (Emily Pike), and physiology/validation (Georgia Chen)SR&ED compliance: Document R&D progress via daily scrum notes for Canadian grant accountingTechnical leadership: Mentor junior interns or co-op hires on ML/backend systems as the team grows
