Agri-Weather Predictor Dashboard

The inspiration for the Agri-Weather Predictor Dashboard came from recognizing how heavily farmers and agricultural planners depend on unpredictable weather patterns. With increasing climate variability, we wanted to create a tool that could transform raw environmental data into meaningful insights for crop yield forecasting. The dashboard provides users with visualizations and predictive analytics that estimate agricultural yield based on historical weather conditions, seasonal trends, and key climate variables. By combining data analysis with an interactive interface, it empowers users to make data-driven decisions about planting, harvesting, and resource allocation. Ultimately, the project bridges the gap between complex weather datasets and practical agricultural planning.

To build the dashboard, we developed a machine learning workflow using XGBoost with a Quantile Regression model within a Jupyter Notebook environment, cleaning and preprocessing historical weather and yield data before training predictive models. We engineered relevant features such as temperature averages, rainfall totals, and seasonal indicators, then evaluated model performance to ensure reliable forecasts. The backend prediction logic was integrated into a user-friendly dashboard interface, allowing users to input conditions and instantly receive projected yield outcomes. Throughout development, we emphasized modular code structure, reproducibility, and clear visualizations so that both technical and non-technical users could interpret the results effectively.

Along the way, we encountered challenges including inconsistent datasets, missing weather records, and model overfitting during early experimentation. Fine-tuning hyperparameters and selecting the most impactful features required multiple iterations and careful validation. Despite these obstacles, we are proud of creating a working predictive system that meaningfully connects climate data to agricultural outcomes. The project strengthened our understanding of data preprocessing, model evaluation, and the importance of clean visualization design. Moving forward, we plan to expand the Agri-Weather Predictor Dashboard by incorporating real-time weather APIs, supporting multiple crop types, and improving model accuracy through additional data sources and advanced machine learning techniques.