Rapid urbanization has intensified the challenges of municipal solid waste management, particularly in ensuring proper waste segregation at the source. This paper presents SwatchhAI, an Android-based smart waste segregation and collection system that leverages Convolutional Neural Networks (CNN) and cloud-based services to automate waste classification without the need for physical sensors or IoT hardware. The proposed system allows users to capture or upload images of waste materials through a Flutter-based mobile application. These images are processed by a CNN model deployed using TensorFlow Lite, which classifies waste into three categories: dry, wet, and hazardous. The system integrates Firebase for secure user authentication, real-time database management, and cloud storage. Additional features include GPS-based waste pickup requests, real-time tracking, push notifications, and a reward-based incentive mechanism. Experimental evaluations demonstrate an overall classification accuracy of 85–92%, confirming the viability of the proposed approach. The system contributes to smart city initiatives by promoting responsible waste disposal and reducing environmental pollution through intelligent, scalable, and cost-effective technology.