Data Analysis Automation Scope
Budget: $30 – $250 CAD
I realised my original listing didn’t spell out the true aim of the work. The core goal is to automate our data-analysis process from raw extraction right through to the finished reports we share each week.
At present we manually pull figures from MySQL tables and a handful of CSV exports, clean everything in Excel, then rebuild the dashboards in Power BI. I want a single, repeatable pipeline that:
• connects to those sources,
• performs the cleansing, joins and aggregations, and
• produces the same visuals or equivalent PDF/HTML summaries—either on a schedule or at the click of a button.
Python, R, or another open-source stack is fine so long as it can run on Windows Server 2019 and won’t lock us into expensive licences.
Deliverables
• Scripted ETL workflow with clear, commented code.
• Automated report generation matching the current layout.
• Walk-through documentation and a brief hand-over call so my in-house analyst can keep it running.
Acceptance criteria
• End-to-end job finishes in under ten minutes on a 2 GB test dataset.
• No manual steps once the schedule is set.
• All dependencies documented and easily installable.
If you have a more elegant toolset or architecture that hits these same outcomes, feel free to outline it—flexibility is welcome as long as the end goal of seamless, hands-off data analysis is achieved.
At present we manually pull figures from MySQL tables and a handful of CSV exports, clean everything in Excel, then rebuild the dashboards in Power BI. I want a single, repeatable pipeline that:
• connects to those sources,
• performs the cleansing, joins and aggregations, and
• produces the same visuals or equivalent PDF/HTML summaries—either on a schedule or at the click of a button.
Python, R, or another open-source stack is fine so long as it can run on Windows Server 2019 and won’t lock us into expensive licences.
Deliverables
• Scripted ETL workflow with clear, commented code.
• Automated report generation matching the current layout.
• Walk-through documentation and a brief hand-over call so my in-house analyst can keep it running.
Acceptance criteria
• End-to-end job finishes in under ten minutes on a 2 GB test dataset.
• No manual steps once the schedule is set.
• All dependencies documented and easily installable.
If you have a more elegant toolset or architecture that hits these same outcomes, feel free to outline it—flexibility is welcome as long as the end goal of seamless, hands-off data analysis is achieved.