The research introduces an innovative supply chain management system tailored for agriculture, aiming to enhance operational efficiency and transparency through cutting-edge technologies such as deep learning and blockchain. The system comprises four modules, each leveraging machine learning algorithms to optimize various agricultural processes. The first module utilizes LSTM models to predict crop yields and diseases, empowering farmers with valuable insights into crop health and potential yields. The second module focuses on inventory management optimization using DQN algorithms, enabling farmers to efficiently utilize resources and minimize waste. The third module employs GNN technology to optimize routes for accessing industry-specific points of interest, simplifying farmers' access to relevant information and resources. Lastly, the fourth module addresses significant challenges in agriculture, including crop demand, rotation planning, market trends, pricing strategies, and market identification, employing a DBN model. This study showcases the transformative power of machine learning algorithms in revolutionizing productivity and profitability within the agriculture sector, offering actionable insights and optimization tools to farmers.