Part-Time Sales Forecast Analyst
Budget / SalaryHourly project
TypeFreelance project
LocationRemote
Posted2 hours ago
I’m looking to bring in a part-time data analyst who can apply predictive analysis techniques to my customer data so I can forecast monthly and quarterly sales with greater accuracy. The raw information is already cleaned and stored in CSV and a Postgres database; what I need now is someone who can turn that history into forward-looking insights.
You’ll have flexible hours—as long as reliable forecasts and clear, actionable summaries arrive on the agreed schedule. I’m comfortable with you working in Python (pandas, scikit-learn, Prophet) or R (tidyverse, caret, forecast); SQL proficiency is essential for grabbing the right slices of data yourself.
Deliverables
• A well-documented forecasting model (code + environment file)
• A brief report that explains the drivers influencing the projections and the confidence intervals for each period
• A short Loom or Zoom walkthrough so I can understand how to update or retrain the model on my own
Acceptance Criteria
• The model must outperform a naïve baseline (e.g., last-period sales) by a statistically significant margin on a held-out test set
• Forecasts cover the next 3, 6, and 12 months and include 95 % confidence bounds
• All notebooks/scripts run end-to-end on my machine without modification
If this sounds like a good fit, please tell me which forecasting libraries you prefer and share a quick example of a sales or demand model you’ve built.
You’ll have flexible hours—as long as reliable forecasts and clear, actionable summaries arrive on the agreed schedule. I’m comfortable with you working in Python (pandas, scikit-learn, Prophet) or R (tidyverse, caret, forecast); SQL proficiency is essential for grabbing the right slices of data yourself.
Deliverables
• A well-documented forecasting model (code + environment file)
• A brief report that explains the drivers influencing the projections and the confidence intervals for each period
• A short Loom or Zoom walkthrough so I can understand how to update or retrain the model on my own
Acceptance Criteria
• The model must outperform a naïve baseline (e.g., last-period sales) by a statistically significant margin on a held-out test set
• Forecasts cover the next 3, 6, and 12 months and include 95 % confidence bounds
• All notebooks/scripts run end-to-end on my machine without modification
If this sounds like a good fit, please tell me which forecasting libraries you prefer and share a quick example of a sales or demand model you’ve built.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.