Python Predictive Sales Model

I have several years of historical sales records spread across CSV files, Excel workbooks, and a database. I want you to turn that mixed-source information into a clean, consistent dataset, explore it for trends and seasonality, and then build a machine-learning model that can reliably project future sales.

Data preparation must cover every stage—handling missing values, normalising numeric fields, and performing any transformations needed to feed the algorithms. Once the data is tidy, run an exploratory analysis so I can clearly see the patterns that drive revenue, especially how previous sales, marketing spend, and seasonal shifts interact.

For modelling, I am open to linear regression, Random Forest, ARIMA, or another regression-based technique you can justify in the report. The core objective is accuracy and interpretability, so include performance metrics that show why the chosen approach works best.

Deliverables
• The fully pre-processed dataset in a convenient format
• Well-commented Python scripts (pandas, scikit-learn or statsmodels, matplotlib/seaborn, etc.)
• Forecast plots that visualise at least the next forecast horizon
• A concise report explaining preprocessing choices, EDA insights, model selection, evaluation results, and how to update the model with new data

If anything in the brief needs clarification, let me know early so we keep the project moving smoothly.