Chronic AI

Chronic AI

Summary

Intelligent Health Insight Engine (2025) Personal project built for chronic illness patients, integrating HealthKit, weather + gait + biometric data for intelligent flare-up detection and chronic pain management.

Problem

Millions of people with chronic illness struggle to understand what triggers their flare-ups and manually tracking symptoms is exhausting. Most apps offer passive logging, not real insight.

Goal

Build proactive system that chronic pain patients more control over their health and respond using real-world data

Role

Design and system architecture Built MVP using SwiftUI, and CoreML Designed data pipelines using Apple HealthKit, WeatherKit, and gait analysis Developed model evaluation criteria and UX for insight deliver

Process

Mapped user pain points with chronic illness experts and patients Scoped MVP to fuse 3 data sources: weather, biometrics,(HRV, sleep, HR), and gait Designed intermediate-fusion ML architecture for pattern detection Created insight cards(e.g., ‘Consider taking it easy today’) based on DeepHeart-style thresholding Built analyst dashboard to validate patterns before surfacing them to users

Outcomes

MVP in use for daily insights Gained signal on HRV + weather correlation with flare-ups Demo-ready prototype showing proactive recommendation tied to user-specific triggers Send health updates to trusted people as needed

Lessons

Simpler insights beat flashy visualizations when users are in pain Timing matters: batch processing worked better than real-time Model transparency and explainability boosted user trust