The agricultural sector is a vital part of the Indian economy, comprising 18.2% of India’s GDP and representing approximately 44% of the total labour force. However, one of the biggest problems faced is the loss of crop yield, especially among farms using traditional methods of farming that lack the technological means to predict and maximise their potential yield. The problem is further compounded by farmers often being unaware of which crops are suitable, given conditions that are specific to individual farmers or parcels of land. This research paper focuses on maximising crop yield by helping farmers choose a suitable crop in Rajasthan, one of the largest Indian states by land mass and population, where over 54% of citizens depend on agriculture as a primary source of income. The data used throughout this paper are publicly accessible and are taken from multiple official Indian government sources. Using these data, the paper incorporates exploratory data analysis to identify key variables such as soil nutrient levels, rainfall, and temperature that influence crop performance. Furthermore, the paper aims to lay out the groundwork for building a crop yield prediction and, primarily, a crop recommendation model that is easily accessible and simple to understand. This is implemented using a transparent linear regression baseline and a decision-tree-based ensemble approach, specifically Random Forest.