SuperKart is an end-to-end machine learning sales forecasting system designed to predict retail demand and support data-driven planning. I developed the pipeline using Python and modern machine learning techniques.

First, I processed historical sales data and engineered predictive features that capture seasonal trends, product behavior, and demand patterns. Then, I trained gradient boosting models to generate accurate forecasts.

In addition, I built a deployment pipeline using Docker and a Flask API so the model can run as a scalable forecasting service. The project also includes a Streamlit dashboard that allows users to visualize predictions and explore sales trends interactively.

Overall, the SuperKart system demonstrates how machine learning can transform raw retail data into actionable business insights. It shows my ability to design, train, and deploy real-world predictive analytics systems.