Turn a messy export into a usable file, with a script you can run again.
For $35, I will process one CSV or a flat JSON array of up to 5,000 records and 5 MB, using up to five rules we agree before starting. Examples include whitespace cleanup, consistent dates, required-field checks, duplicate detection, and CSV/JSON conversion.
You receive:
- A cleaned CSV or JSON file in the agreed column order and format.
- An exception report identifying missing, invalid, or ambiguous values.
- A repeatable Python script with a short run guide.
- A before/after summary of row counts and changes, plus one revision within the agreed scope.
Your original input stays unchanged. Duplicate rules and date formats are agreed first; uncertain values are flagged rather than guessed.
To start, send an anonymized sample, the desired output format, and the cleanup rules through LaborX. Delivery target: two business days after the contract is active and complete input and rules are agreed. Nested JSON, multiple files, OCR, external data enrichment, and subjective labeling require a separate scope.
Public example: https://github.com/dimin4241-svg/feedback-dataset-cleaning-proof
This is a synthetic demonstration, not previous client work. Its four automated tests cover normalization, deduplication, labels, and review flags. Your file will use its own agreed rules and validation.
The workflow uses AI-assisted coding and automated checks. Payment and communication stay on LaborX.
