Date of Award

6-4-2026

Document Type

Thesis

Publisher

Santa Clara : Santa Clara University, 2026

Department

Computer Science and Engineering

First Advisor

Sean Choi

Abstract

Insufficient sleep is a widespread public health concern, yet many consumer sleep tools still emphasize retrospective dashboards over personalized, actionable guidance. Platforms such as Apple Health, Fitbit, Whoop, and Oura collect detailed biometric and sleep data, but users are often left to interpret trends and decide what to change on their own.

SleepFocus was built to address that gap by turning wearable sleep data into clearer feedback, personalized recommendations, and healthier sleep routines. SleepFocus is a native iOS application that integrates with Apple HealthKit to collect sleep-stage data and biometric streams such as heart rate, heart rate variability, respiratory rate, wrist temperature, and environmental noise from HealthKit-compatible devices. The data is sent to a backend pipeline that cleans and stores the measurements, computes sleep scores, builds circadian sleep-timing profiles, and generates model-backed insights.

The Smart Alarm feature uses the circadian profile and sleep-cycle guidance to recommend wake times within the user’s allowed wake window. Instead of relying only on a fixed alarm time, SleepFocus favors cycle-aligned wake opportunities that are less likely to occur during deeper sleep, with the goal of reducing sleep inertia. The system is supported by a Node.js backend, MongoDB for persistent data storage, Redis for asynchronous pipeline jobs, and a Python modeling service for scoring, circadian analysis, and wake-time recommendations.

SleepFocus demonstrates how wearable biometric data, circadian modeling, and large language model reasoning can be combined into a sleep wellness system that moves beyond passive tracking toward personalized, actionable support.

Download

Share

COinS