Data Validation

Data validation is the process of checking that incoming data meets defined rules and constraints before it is accepted into a system or pipeline. Validation ensures data is complete, correctly formatted, within expected ranges, and internally consistent.

Validation rules might include: price fields must be positive numbers, listing dates must be in the past, required fields must not be null, and status values must belong to an allowed set. Validation catches errors at ingestion - before they propagate into reports, models, or user-facing features.

Validation is distinct from data cleaning: validation checks whether data conforms to rules and rejects or flags what does not, while cleaning actively transforms non-conforming data. Both are necessary in a production-grade pipeline - validation as the gatekeeper and cleaning as the repair layer.