A machine learning project that predicts soil fertility levels using soil nutrient data, implemented end-to-end in a Jupyter Notebook.

Time-series modeling pipeline to predict cotton growth stage (GSTD) and generate an interpretable risk score using lag features, LightGBM, and SHAP

Interactive Streamlit dashboard simulating agricultural yield interventions in Ethiopia using predictive modeling, economic impact analysis, and regional insights from synthetic 2000–2023 crop data.