Data Science Internship Tasks Help -- 2

I’m midway through my internship at Devnest and have been handed seven separate challenges that must be wrapped up before the final review. The scope spans the full data pipeline: first cleaning messy raw files, then exploring them, and finally building and evaluating machine-learning models.

Everything lives in Python and SQL. Day-to-day I lean on Pandas for wrangling, NumPy/SciPy for statistical work, and Plotly or Seaborn for quick visuals. Dashboards need to be surfaced either in Tableau or Power BI, and a couple of interactive prototypes are earmarked for Streamlit or D3.js. One task dives into time-series forecasting with Prophet or ARIMA, while the others involve clustering, association-rule mining, anomaly detection, and a graph-analytics proof of concept.

What I need from you is production-ready code, cleanly commented notebooks or .py scripts, the corresponding SQL queries, and polished visuals that can drop straight into a presentation. Each task should close with a short markdown report explaining the approach, assumptions, validation metrics, and next steps. Code must run end-to-end on fresh data without tweaks; models should beat the provided baselines, and any dashboard links should be shareable and interactive.

If you’ve shipped similar projects before and are comfortable switching between data cleaning, exploratory analysis, and model development in the same sprint, you’ll fit right in.