Kinevra

Inspiration

Kinevra is an assistive research prototype designed for movement monitoring between physiotherapy visits. It is not a medical device and does not provide medical diagnoses. Users are instructed to cease exercise immediately if discomfort or pain occurs.

What it does

Kinevra is an agentic computer vision framework built to monitor movement during home rehabilitation exercises. Powered by OpenCV 5 DNN pose estimation and Amazon Bedrock via LangGraph, Kinevra continuously measures biomechanical features, detects execution deviations, dynamically triggers secondary computer vision re-perception tools, and escalates flagged sessions to human clinical reviewers.

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for Kinevra

Built With

• amazon-web-services
• opencv
• python