SPYN — AI Posture Monitor & Exercise Coach

Built at PennApps XXVI (UPenn) in 48 hours — a computer vision AI posture monitor giving real-time corrective feedback during work and exercise.

Overview

SPYN is a real-time posture monitoring and exercise coaching system built at PennApps XXVI at the University of Pennsylvania. The system uses pose estimation to analyze a user's posture, providing corrective feedback to prevent pain and improve performance. Conceptualized and built entirely in 48 hours.

Role

Lead Developer — Conceptualized the system, built the CV pipeline, designed the real-time feedback engine, and delivered a working demo within the 48-hour hackathon window.

Tech Stack

•Computer Vision (MediaPipe Pose)
•Python
•OpenCV
•Real-time Feedback Engine
•Web Interface
•AI / Pose Estimation

Key Challenges

•Building a complete, functional product in 48 hours
•Achieving accurate real-time pose estimation on consumer hardware
•Designing actionable posture feedback that users can immediately apply
•Handling varied environments and lighting for reliable detection

Results

•Successfully built and demonstrated functional posture monitor at PennApps XXVI (UPenn)
•Real-time pose estimation with immediate corrective feedback
•Complete end-to-end system from CV pipeline to user interface in 48 hours

Interested in learning more or collaborating on similar projects?

Get In Touch