Managing the waste serves crucial for ecological balance, community well-being and economic productivity. The research work examines the assimilation of connected devices and advanced machine learning models to optimize the classification and management of solid waste in Indian cities. Utilizing advanced machine learning algorithms including convolutional neural networks and deep learning models are used in analyzing and classifying the waste producing high accuracy. The proposed system overcomes the disadvantages of manual sorting and classifies the images of solid waste into three labels such as organic, recyclable and non-recyclable wastes. The solid wastes after mere classification are processed separately were organic waste can be used as fertilizers, recyclable waste are recycled used in reprocessing industries whereas the non-recyclable waste can be processed by incineration and pyrolysis technologies which converts the non-recyclable waste to energy, thus reducing land fill use and generating electricity, such that the environmental impact can be minimized. The particulars of the waste is sent to through IoT devices to the mail for decisive purpose. The initial objective was to develop a Custom CNN model which produced the accuracy of 77% when compared with the hybrid algorithms, whereas Customized InceptionNet model when incorporated with IoT produced an accuracy of 94%. This paper puts forth the challenges in the implementation, usage of hybrid algorithms, environmental benefits and future research opportunities, highlighting the impact of IoT and advanced machine learning towards waste management.