METADATA-DRIVEN TIERED STORAGE

A metadata-driven file system prioritizes the distribution of files across different storage devices based on arbitrary file metadata matching to different file usage scenarios defined within one or more workflows or taxonomies. The file metadata may be tagged with different key-value pairs from the defined workflows or taxonomies. Upon receiving a file or file access request, the metadata-driven file system matches the file metadata to a key-value pair combination of the taxonomy, and stores the file to a storage device that is selected using a storage rule that is defined for the matching key-value pair combination. The taxonomy may be defined with other programmatic file operations that the metadata-driven file system may perform against files with matching metadata.