Demand forecast model
A forecasting model to reduce stockouts and overstock using historical sales data.
Problem
This project looks at how to estimate future demand to reduce two common problems: stockouts and excess inventory. The goal is to turn sales history into a tool that supports purchasing, production, and planning decisions.
Data source
Practice project using public or simulated historical sales and demand data. The dataset includes dates, products, and units sold, plus — where available — variables like promotions, seasonality, or inventory levels.
Analysis
The analysis covers demand evolution, trends, seasonality, and variation by product or period. Historical behavior will be compared and a forecast model will be built, evaluating its error to determine how useful it is for planning.
Solution
The output is a demand forecast alongside a view comparing actual vs. estimated demand. Based on the results, recommendations will be proposed for inventory planning and for prioritizing products at higher risk of stockout or overstock.
Results & impact
Practice project. Impact will be evaluated through the model's metrics and the usefulness of the recommendations generated. No real reductions in inventory, cost, or waste will be attributed without implementation data.