Project Description
I’m sitting on several datasets—a mix of tidy spreadsheets and raw text logs—and I need them transformed into business-ready insights with reliable predictive power. Your first task will be to clean and consolidate this structured and unstructured data in Python using Pandas / NumPy and, where helpful, SQL queries. Once the data pipelines are solid, I’d like an exploratory analysis that highlights key patterns or anomalies before we move into supervised models that forecast the metrics I care about most.
For visual communication, every important finding should flow into an interactive Power BI dashboard. I lean on these dashboards to brief non-technical stakeholders, so clarity and drill-down capability matter as much as aesthetic polish.
Deliverables
• Cleaned, documented dataset(s)
• Jupyter notebook (or .py script) showing EDA and model development
• Trained predictive model with performance metrics and reproducible code
• Power BI file with interactive visuals and written insights/commentary
Acceptance criteria
• Model performance ≥ the baseline we set together during kickoff
• Code runs end-to-end on my machine with a single command
• All visuals refresh correctly when data is updated
If you enjoy iterative, long-term collaborations and have a solid grasp of Python, basic machine-learning algorithms, and polished Power BI storytelling, this should be a straightforward, rewarding partnership.