Soil Organic Carbon (SOC) is an indicator of soil fertility and plays an important role in mitigating climate change due to its potential to sequester atmospheric carbon. Laboratory SOC analysis methods are resource-intensive and impractical for large-scale applications. Therefore, there is a need to explore new methods to estimate this element. In this context, a Hackathon was held at the European Space Agency Symposium on Earth Observation for Soil Protection and Restoration, in which the main focus was the development of artificial intelligence models in SOC prediction using remote sensing data. This paper presents the proposed solution, which used Recursive Feature Elimination with Cross-Validation to select the most relevant features and the Extreme Gradient Boosting algorithm to estimate SOC, which a Lazy Regressor indicated as the best regressor for the work in the case. The developed solution achieved a Root Mean Square Error of 0.43354 on the private dataset and won second place in the Hackathon. This result demonstrates the potential of using Machine Learning models and Remote Sensing data in estimating SOC, making obtaining data for large areas at low costs easier, contributing to sustainable land management and climate action strategies.