Fully unit tested utility functions for data engineering. Python 3 only.
Conforms pandas to "correct" datatypes to ensure data in/out using CSV, JSONL and Parquet is read the same (using arrow).
Docker image used to automatically validate data
