PULIDs are an identifier encoding scheme that builds upon the excellent Universally Unique Lexicographically Sortable Identifier scheme. Prefixes should maintain compatibility with base32. The library uses the following alphabet: 0123456789ABCDEFGHJKMNPQRSTVWXYZ. If constrained to 2 characters for encoding the entity type, the number of types of entities that can be referenced with this scheme is 2^32 (1024). Though keep in mind that obvious encodings of table names will start to overlap far sooner.