Pandera is an open-source Python library for validating dataframes. Data scientists and engineers define schemas and checks, then validate data at the entry and exit points of a pipeline. Pandera can also infer schemas from clean data and generate synthetic test data. The library supports pandas, Polars and PySpark, among others, is released under the MIT licence and is installed as an ordinary Python package from PyPI or conda-forge. It therefore runs inside your own environment rather than as a hosted service.
Pandera began as a community project and has been maintained by Union.ai, a US company, since 2022. Because the source code is open, you can inspect it, fork it and pin a specific version, and the MIT licence allows this even if the direction of the project changes. For a European organisation, the dependency therefore lies mainly in the future direction of an open-source project steered by a US company, not in where data is hosted, since validation happens locally within your own code and infrastructure.