Numerous obstacles endanger the livelihoods of smallholder farmers and global food security. Crop output losses due to plant diseases alone are estimated to be between 15% to 25% every year. However, farmers' capacity to recognize and treat illnesses in a timely manner is hampered by things like a lack of agronomic expertise, limited access to actionable insights, and information gaps. The work proposes a comprehensive end-to-end web application that uses machine learning models to help farmers with configurable expert advice, smart farm management, and automatic disease detection. Thus, enabling farmers to identify the infection using their smart phones and obtain crop - fertilizer recommendations for the crops to be treated in the farmer's area based on historical data analysis of yield-influencing elements such as geography, weather, and soil. The proposed system will employ Generative Artificial Intelligence, such as ChatGPT's natural language processing and generation capabilities, to engage in conversational interactions with farmers. The work achieves close to 95% accuracy after utilizing 30 epochs in the area of disease detection and 97% accurate models are achieved for fertilizer outputs. These features along with it's high accuracy can assist farmers in making informed decisions, utilising the data obtained to reduce crop output losses.