Rice is a critical staple for food security in Mauritania, but its production faces challenges due to climate variability and resource constraints. This study proposes an integrated framework that combines artificial intelligence (AI) and the Internet of Things (IoT) technologies to improve rice yield prediction and precision agriculture. Leveraging datasets on rice yields (1961–2020), agricultural value added (1960–2023), rice prices, and agricultural indicators such as fertilizer use and rural population, we employed ensemble machine learning models (Random Forest, XGBoost) and Long Short-Term Memory (LSTM) networks. IoT data simulated from historical climate records due to the absence of real-time sensors provide temperature and humidity information, improving model accuracy, although this simulation limits real-world applicability. Analysis of rice price data revealed a weak correlation (0.081325) between retail and wholesale prices, indicating limited market integration. The preliminary results suggest a high prediction accuracy (R-squared > 0.85), with Random Forest outperforming the baseline models. Practically, this framework enables Mauritanian farmers to optimize planting schedules and resource allocation, enhancing climate resilience and potentially increasing yields by 10–15%. Future work will explore real-time IoT deployments and scalability in Sub-Saharan Africa.