Machine Learning Engineer Needed for Sales Data Analysis – Summer Plan
Budget: $10 – $30 USD
My complete sales history is already in Excel/CSV, but for trend discovery I need you to drill down on the last twelve months, then roll those findings forward into a clear summer-season action plan.
The raw file will be shared as soon as the project starts. Once you have it, please clean, explore, and model the data in Python (pandas, scikit-learn, or similar). Visual explanations through matplotlib or seaborn charts are welcome, though the centerpiece must remain a concise, manager-friendly report.
Deliverables
• Short report (PDF or slide deck) summarising:
– Key sales trends & purchasing patterns from the past year
– Recommended product collections most likely to resonate this summer
– Offer/discount structures that align with those patterns
– Social-media-focused marketing ideas backed by the data
– A numerical forecast of expected summer sales, with underlying assumptions
I will validate the work by checking that:
1. All figures in the report can be traced back to the code or notebook you provide.
2. Forecast error metrics (e.g., MAE or MAPE) are shown.
3. Recommendations are specific, actionable, and linked to the discovered patterns.
Keep the language of the report practical; I want insights I can act on the day after delivery.
The raw file will be shared as soon as the project starts. Once you have it, please clean, explore, and model the data in Python (pandas, scikit-learn, or similar). Visual explanations through matplotlib or seaborn charts are welcome, though the centerpiece must remain a concise, manager-friendly report.
Deliverables
• Short report (PDF or slide deck) summarising:
– Key sales trends & purchasing patterns from the past year
– Recommended product collections most likely to resonate this summer
– Offer/discount structures that align with those patterns
– Social-media-focused marketing ideas backed by the data
– A numerical forecast of expected summer sales, with underlying assumptions
I will validate the work by checking that:
1. All figures in the report can be traced back to the code or notebook you provide.
2. Forecast error metrics (e.g., MAE or MAPE) are shown.
3. Recommendations are specific, actionable, and linked to the discovered patterns.
Keep the language of the report practical; I want insights I can act on the day after delivery.