This paper presents VisionAid, an AI-assisted platform built to help blind and low-vision users move about more freely and stay aware of what is around them. Instead of piecemeal apps that handle single tasks, VisionAid works as a hands-free, all-in-one companion. The system blends live hazard spotting, scene breakdown, voice-guided prompts, and plain-language help. Key components include a YOLO model for obstacle tracking, a BLIP model for scene reading, and LLaMA 2 for conversational replies. User comfort grows through a neural voice-cloning unit that speaks in tones the person already trusts. Built on a modular cloud back end, the service keeps lag low, scales easily, and delivers smooth talk-based navigation. Early trials in homes, streets, and transit hubs point to big gains in everyday mobility, independence, and dignity for users with vision loss.