Portable Format for Analytics (PFA) is an open standard for representing statistical and machine learning models, along with their associated data transformations, in a portable and executable format. Unlike PMML, which focuses on predefined model structures, PFA offers greater flexibility by expressing models, pre-processing, and post-processing as composable functions using JSON or YAML. This makes PFA better suited for defining complete analytic workflows that can be deployed consistently across different environments, enabling dynamic, scalable, and vendor-independent execution of complex AI and ML pipelines.