Welcome to Sparkle ✨! To learn more about us, feel free to peruse the sections below. Or if you're eager to jump right in, you can visit our website sparklemed.com and experience Sparkle firsthand.

Sparkle grew out of an Apache Spark based research project, which used distributed machine learning and wearable technology to detect low pill counts in prescription containers, with the goal of automating the delivery of refillment medication. It has since grown into a comprehensive medication adherence platform for doctors and patients alike that seeks to provide an intelligent solution to the issue of medication adherence.

Why? It's estimated that in developed countries approximately 50% of patients with chronic illnesses do not take their medication as prescribed. In the United States alone, hospitalizations due to medication non-adherence cost an estimated 289 billion dollars annually.

Sparkle's mission is to eliminate medication non-adherence by providing a comprehensive multi-platform medication monitoring system.

Our system joins patients and doctors in a single feedback loop in which:

• Doctors upload patient medication information to our internal online portal, with end-to-end encryption of patient data.
• Patient prescription schedules are automatically propagated to our mobile iOS and watchOS applications.
• Patients are automatically notified on their AppleWatch when it's time to take their medication. No configuration required.
• Pill intakes are recorded via AppleWatch sensors and sent to our web API, where intakes are verified and the number of pills remaining is estimated via a Gradient Boosted model.
• Our online portal is updated with patient intakes. Doctors can view the adherence rates of individual patients and are immediately notified of any non-adhering patients so that they can intervene and help patients resume their prescribed course of treatment.

Sparkle provides a comprehensive medication monitoring system that is unlike anything on the market to date. Whereas other medication adherence applications require users to manually input a detailed intake schedule for each prescription they take, Sparkle reverses this paradigm and leverages the patient information that doctors must already record in online medical record systems to ease the burden of medication adherence for patients at no added cost to medical professionals.

Want to learn more more about Sparkle? Watch the video below for a live demo of how our iOS, watchOS, and web applications work together, or visit our website to experience Sparkle firsthand.

Sparkle was created and developed with ❤️ over the course of seven weeks as part of the Product Analytics class at the University of San Francisco's Masters in Data Science Program.

Want to learn more about our team? Feel free to reach out using the links below!

Andy Cheon | Stephanie Jung | Collin Prather | Matthew Sarmiento | Kevin Wong