KNIME-Python Mixed Data Cleanup

I have a one day assignment centred exclusively on data cleaning and preprocessing. The raw sources arrive in mixed formats—numbers, categorical fields, free-text comments—so I need a specialist who feels at home moving across KNIME, SQL, Python and, once the dust has settled, feeding the tidied tables into Power BI for quick checks.

Here is what success looks like to me:
• An end-to-end KNIME workflow (or equivalent Python script, where it makes more sense) that imports the files and database tables, handles missing values, normalises data types, and flags outliers without hard-coding paths.
• Reusable SQL snippets for in-base cleansing steps.
• A final, well-documented output table or file that drops directly into Power BI with no additional manipulation needed.

Please highlight similar mixed-data cleaning projects you have tackled; I am choosing primarily on demonstrated experience, not on elaborate proposals. A concise note explaining where you have used KNIME or Python for preprocessing and how you validated the results will help me make a quick decision.