Kaushalendra Pratap Singh
AI Engineer
Demand Forecasting System
Time-series prediction for inventory planning
Problem
Inventory planning in retail lacks predictive accuracy, leading to either stockouts during high-demand periods or excess inventory during slow seasons.
Approach
Built a time-series forecasting system using Prophet that models trend, seasonality, and holiday effects on historical sales data. The system decomposes the signal to provide interpretable predictions.
Outcome
Delivered actionable forecasts supporting inventory planning and demand optimization.
Tech
PythonProphetMLflowPandasFastAPI