I have a historical daily sales dataset ready to share and I want a reliable forecasting model built in Python within the next three days. My focus is strictly on time-series techniques; I am happy with either ARIMA or Facebook Prophet (or a thoughtful combination if you see value in blending them) and I’m open to your recommendations on parameter tuning, seasonality handling, and any necessary data transformations. Here is what I need from you: • Clean, prep, and explore the data so that any missing values, outliers, or calendar effects are handled transparently. • Build the forecasting pipeline in reproducible Python code (Jupyter Notebook or .py script). • Generate point forecasts and confidence intervals for a horizon we will decide together once you see the file. • Visualise actual vs. predicted values and provide a concise summary explaining model choice, diagnostics, and performance metrics (MAE / RMSE). • Package everything—code, plots, and brief read-me—so I can rerun the model on fresh data without issues. The dataset will be provided immediately after project start, and I’m available to answer questions quickly so we stay on schedule. If you are comfortable working with daily-frequency data and have hands-on experience with ARIMA or Prophet, I’d love to see your approach.