Rice production faces constant threat diseases. These diseases can devastate rice crops, leading to reduced yields, economic hardship for farmers, and threats to food security in rice-dependent regions. Current methods of paddy disease detection rely on visual inspection, which is subjective, expertise-dependent, and time-consuming. Additionally, traditional fertilizer application methods often lack disease specificity, leading to over- or under-fertilization. Our paper addresses these challenges by leveraging deep learning techniques to develop a comprehensive system for paddy disease management and fertilizer optimization. It identifies various paddy diseases from images with high accuracy, enabling early intervention and improved treatment strategies. Recommend appropriate fertilizers based on the identified disease, promoting optimized resource allocation and potentially higher yields. Our system employs Dense Net, a Convolutional Neural Network (CNN), to empower farmers with a tool capable of accurately identifying both paddy plant varieties and diseases from image inputs. By leveraging transfer learning and hyper parameter tuning like learning rate and batch size, enabling fine-tuning for optimal model accuracy, this model is trained on a diverse dataset encompassing various paddy plants and disease manifestations, ensuring robust performance in real-world scenarios. This proposed system is capable of identifying diseases and various plant varieties within paddy crops. This streamlined approach empowers farmers and also to new people in this field, to promptly address concerns and optimize crop health, thereby fostering heightened productivity and sustainability in paddy cultivation practices.