Otutu11/Carbon-Sequestration-Potential-Mapping-with-AI

The project is explicitly defined as a demonstration and teaching tool rather than a production-grade utility. With a defensibility score of 2, it functions primarily as a portfolio piece or a pedagogical resource. It relies on synthetic data rather than real-world satellite or soil datasets, which removes the 'data gravity' moat essential for geospatial AI. From a technical standpoint, it uses standard scikit-learn Random Forest implementations, which are commodity tools in the data science community. While it has 39 stars—indicating some visibility—the lack of forks and zero current velocity suggests it is a static repository rather than a living project. Competitively, it does not challenge professional carbon accounting platforms like Pachama or Sylvera, which utilize high-resolution LiDAR, SAR, and specialized computer vision architectures. Frontier labs are unlikely to compete directly as this is a niche environmental application, but any competent data scientist could replicate this workflow in hours, making its displacement horizon very short.