Mobile App Development
Shut Eye
Sleep tracking app with Python ML algorithms analysing patterns for personalised insights.
Industry
Health & Wellness
Platform
Mobile (iOS & Android)
Year
2023
/ Overview
About the project.
Shut Eye is a comprehensive sleep tracking and improvement app with millions of downloads. As a Backend Developer, developed machine learning algorithms using Python to analyse sleep patterns and improve tracking accuracy — delivering personalised sleep stage insights to users each morning.
Visit live siteAccurate
Sleep stage classification
Millions
App downloads
Personalised
Nightly sleep insights
Efficient
ML inference performance
/ Challenge
The problem we had to solve.
Developing ML algorithms accurate enough to classify sleep stages from device sensor data, while keeping inference efficient enough to not impact device battery life or interrupt the sleep experience.
/ Solution
How we turned it into a working product.
Built Python machine learning models for sleep stage classification using accelerometer and audio signal inputs, optimised for inference efficiency and deployed as a backend service delivering results to the Flutter frontend.
/ Technology
Stack behind the build.
The tools and platforms used to ship the product, grouped by responsibility.
Mobile
Backend / ML
Python/ Team
The delivery team.
3 specialists contributed across delivery, engineering, design, infrastructure, and QA.
/ Goals
Project Goals
Build ML-powered sleep analysis that gives users genuinely accurate, actionable insights about their sleep — not generic advice, but personalised analysis grounded in each night's real sensor data.
ML Sleep Stage Classification
Developed Python machine learning algorithms that classify deep, light, and REM sleep stages from sensor inputs with high accuracy — giving users detailed, personalised breakdowns each morning.
Efficient Inference Pipeline
Optimised the ML inference pipeline to run as a lightweight backend service, delivering sleep analysis results quickly without requiring heavy on-device computation that would drain battery during the night.