Sales Data Dashboard & Tutorial
Budget / Salary$10–30
TypeFreelance project
LocationRemote
Posted1 hour ago
The project centres on my raw sales data, delivered to you in a mix of CSV files and Excel spreadsheets. I need the full analytical pipeline built and documented so that anyone in my team can reproduce it.
First, consolidate every source in Excel and use Power Query to remove duplicates, fix date and currency formats, and create calculated columns that will later support trend analysis, sales-performance comparisons and customer segmentation. Keep the transformation steps clearly named and ordered so that refreshing future files is one click.
Next, publish the cleaned table to Power BI and design an interactive, professional-looking dashboard. Key visuals should highlight monthly and quarterly trends, compare performance across product lines or regions, and let the user dive into customer segments through slicers or drill-through pages. I appreciate thoughtful colour choices, responsive layouts, and performance-optimised DAX where aggregations are heavy.
After the BI layer is stable, upload the same cleaned dataset to a SQL environment (or load it with Python if you prefer a pure Python workflow). Please include the SQL table definition and a population script, or a Python SQLAlchemy snippet that achieves the same result.
Because part of my team prefers Python, create a concise Jupyter Notebook that walks through the data import, the main transformation logic (mirroring Power Query), and quick visual checks with pandas, seaborn or matplotlib. The notebook serves as a tutorial, so add explanatory markdown cells and well-commented code.
Deliverables
• Excel file with Power Query steps saved
• Power BI .pbix file featuring trend, performance and segmentation visuals
• SQL script (or Python script) that builds and populates the sales table
• Jupyter Notebook tutorial replicating the process in Python
• A brief hand-over document summarising choices, measures, and how to refresh or extend the model
On completion, walk me through each artefact and the reasoning behind key measures so I can maintain and scale the workflow internally.
First, consolidate every source in Excel and use Power Query to remove duplicates, fix date and currency formats, and create calculated columns that will later support trend analysis, sales-performance comparisons and customer segmentation. Keep the transformation steps clearly named and ordered so that refreshing future files is one click.
Next, publish the cleaned table to Power BI and design an interactive, professional-looking dashboard. Key visuals should highlight monthly and quarterly trends, compare performance across product lines or regions, and let the user dive into customer segments through slicers or drill-through pages. I appreciate thoughtful colour choices, responsive layouts, and performance-optimised DAX where aggregations are heavy.
After the BI layer is stable, upload the same cleaned dataset to a SQL environment (or load it with Python if you prefer a pure Python workflow). Please include the SQL table definition and a population script, or a Python SQLAlchemy snippet that achieves the same result.
Because part of my team prefers Python, create a concise Jupyter Notebook that walks through the data import, the main transformation logic (mirroring Power Query), and quick visual checks with pandas, seaborn or matplotlib. The notebook serves as a tutorial, so add explanatory markdown cells and well-commented code.
Deliverables
• Excel file with Power Query steps saved
• Power BI .pbix file featuring trend, performance and segmentation visuals
• SQL script (or Python script) that builds and populates the sales table
• Jupyter Notebook tutorial replicating the process in Python
• A brief hand-over document summarising choices, measures, and how to refresh or extend the model
On completion, walk me through each artefact and the reasoning behind key measures so I can maintain and scale the workflow internally.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.