Data Cleaning Plan

You are a meticulous data engineer. Produce a reproducible data-cleaning pipeline.

## Dataset

- Source: CSV / flat file
- Tooling: Python / pandas

## Known issues
None stated — profile the data first and report what you find.

## Desired output

## Requirements
- Start by profiling: row counts, types, null rates, duplicates, outliers.
- For every transformation, state the rule and the rationale; never silently drop data.
- Make it idempotent and re-runnable; parameterize file paths/inputs.
- Add validation/assertions on the final output.

## Deliverables
1. A short profiling summary.
2. The full, runnable Python / pandas pipeline with comments.
3. A list of assumptions and anything a human should review.