HealthMine

Wearable and mobile devices now make it possible to continuously monitor physiological, behavioral, and environmental signals outside the clinic. But most of these devices operate in isolation: a continuous glucose monitor knows nothing about the wearer’s activity, sleep, or context, and vice versa. The result is a set of disconnected data streams that individually capture only a narrow slice of a person’s health.

HealthMine addresses this gap by integrating disparate streams from consumer and medical-grade devices into a single research platform, so that health-relevant data can be interpreted in context rather than in isolation.

What I built

• Cross-platform mobile clients in Kotlin (Android) and Flutter, with authenticated background synchronization from the Dexcom CGM and Fitbit APIs for continuous, unattended data collection.
• Backend services on Google Cloud Platform written in Go, with Firebase for user management and scalable storage of high-frequency time-series sensor data.
• End-to-end research data pipelines spanning study design, ingestion, preprocessing, feature engineering, and statistical analysis — reused across multiple digital health studies in the lab.

Ongoing user study

Our team is actively recruiting for a user study on this project. We invite anyone over 18 years of age who has diabetes, uses a Dexcom/Freestyle Libre continuous glucose monitor, and owns an Android phone to consider participating. Please email ah-lab@dartmouth.edu to learn more and sign up.