This AI-powered system provides viewpoint-invariant human pose estimation for elderly care applications. It ensures accurate fall detection (>96% accuracy) regardless of camera placement, supports continuous gait analysis for mobility assessment, and delivers real-time alerts at 12 FPS on affordable edge devices. The system reduces setup complexity by working with a single camera from any position. Applications include smart home monitoring, clinical gait assessment, and assisted living support. It improves pose estimation accuracy by 28% under challenging viewpoints, reduces required training data by 60%, and enables privacy-preserving elderly care monitoring.