Built an end-to-end sales forecasting model using Python, Pandas, and Random Forest Regression to predict future sales trends based on historical transaction data.
- Data Preprocessing & Daily Time-Series Aggregation
- Feature Engineering (Lag variables, 7-day rolling trends, day of week indicators)
- Chronological Train-Test Validation
- Business-focused Visualizations
- High sales demand observed in Q4 (November–December) reaching peaks of $6,000–$15,000 daily.
- Recommendation: Increase stock buffer levels by 30% starting late October to mitigate potential stockouts during peak holiday season.