Numerous diseases that can drastically lower yield and quality affect tomato plants, including botrytis grey wilt, grey leaf spot, Verticillium wilt, Fusarium wilt, anthracnose southern blight, septoria leaf spot, early southern blight, bacterial speck, Fusarium wilt, and blossom end rot. For many illnesses to be effectively managed and prevented, early identification and precise diagnosis are essential. In order to sort or grade tomatoes, this study suggests an automated machine learning-based system for detecting plant diseases in tomatoes. Initially, a vast collection of photos showing both healthy tomato plants and a range of frequent illnesses is gathered. Next, using transfer learning of the dataset, a deep Convolutional Neural Network (CNN) architecture is created and trained. When it comes to identifying between healthy and unhealthy tomato plants, the trained model performs with excellent accuracy. Additionally, it can identify the particular disease that is afflicting the sick plants, allowing for more focused management and treatment plans. The suggested technique provides a practical and economical means of identifying diseases in tomato plants early on, which will increase yields and lessen farmers' financial losses. Future research may examine the integration of a robotic platform for autonomous disease monitoring and management, as well as the real-time deployment of the system in agricultural settings.