SuperKart, a retail chain operating supermarkets and food marts across tiered cities, requires accurate sales revenue forecasting to optimize inventory management and inform regional sales strategies

Develop and deploy a production-grade sales forecasting solution for SuperKart, integrating machine learning models with a scalable backend and frontend interface to provide actionable insights

• Data collection and exploratory data analysis (EDA) with visualizations
• Data preprocessing with feature engineering and pipeline creation
• Random Forest and XGBoost model development and hyperparameter tuning
• Model performance comparison and selection
• Model serialization and testing on validation/test sets
• Backend deployment with Flask API and Docker
• Frontend deployment with Streamlit

Python, scikit-learn, Flask, Streamlit, Docker, Graphviz

Deployed Flask API and Streamlit frontend on Hugging Face spaces, optimized Random Forest model (RMSE ~$256, R² ~0.940), actionable business recommendations