This research introduces a new framework using Internet of Things (IoT) and data-warehousing to accurately identify the most suitable crops for specific regions worldwide. The first step of our approach consists of a soil-sampling protocol used with IoT sensors to accurately identify soil properties. Secondly, a data-warehouse is constructed based on web-scraping that collects relevant agricultural information (crops, regions, climate…) sourced from online databases and websites. A specialized tool is developed to convert multimedia content into exploitable datasets. The data-driven model considers a wide range of factors, including soil nutrient levels (NPK), temperature, pH, conductivity, wind sensitivity, and crop-specific requirements. The third step is to filter the data-warehouse content by the soil sample properties obtained from step one to get the most relevant crops for the selected field. A predicted yield is associated to each crop using Machine Learning model trained on historical production data. By providing farmers with informed decisions, our framework aims to enhance agricultural productivity, sustainability, and resilience in the face of resource scarcity. Finally, we show how a large-scale sampling operation can provide a precise map of agricultural lands that allows national crop cultivation and rotation planning. While the experimental study was conducted in Algeria to validate the framework’s effectiveness, its principles remain applicable to diverse agricultural environments worldwide.