OrthIQ
AI‑driven motion intelligence for physical therapy and athletic training
Demo
Overview
OrthIQ is an AI‑driven motion‑intelligence platform created at CalHacks 12.0 (October 2025). The platform pairs a wearable resistance sensor with real‑time biomechanical analysis to guide physical therapy patients through precise movement feedback, merging mobile interaction, data science, and human recovery. Built and shipped production‑grade React Native app for iOS and Android with wearable sensor integration.
Role: Lead Engineer & Co‑Founder
Technical Stack
- •React Native (Expo)
- •Next.js
- •Convex Realtime DB
- •Python
- •C++
- •TypeScript
- •Linear Regression Models
- •Cosine Similarity Scoring
Notable Features & Challenges
- 1Joint‑angle capture via custom sensor mounted to knee or elbow sleeve
- 2Real‑time rep tracking and scoring using cosine similarity
- 3Therapist portal for assigning and monitoring recovery programs
- 4Patient app with haptic feedback, visual reinforcement, and progress reporting
- 5AI assistant estimating range of motion and recovery timeline using regression on user history
Impact
Delivered a fully functional prototype—hardware, mobile app, cloud database, and AI module—within 36 hours at CalHacks 12.0 (October 2025). Achieved reliable motion‑pattern recognition and positive evaluation from multiple orthopedic surgeons at University of Utah Hospital.
Reflection
OrthIQ taught me that technical innovation means little without user empathy. We started with a soldering iron that barely worked, coded through nights of debugging rep‑detection logic, and still delivered something that could help real patients. The project reminded me that great systems often emerge from small frustrations—our own experiences with injury—and that persistence and teamwork can transform discomfort into design.