Named-entity recognition

Find named entities in a text — people, organizations, locations, dates, times, money amounts, percentages, quantities, products, events — with character spans. Returns entities (each with its verbatim text, type, and start/end offsets into the exact submitted text) and types (the distinct types found). Every entity is a span the text actually contains — invented mentions are dropped. Offsets index the submitted text as-is (nothing is stripped or normalized), so they can be applied directly to your copy of the document. Use types to restrict recognition to a subset. Complements sapling_detect_pii: that tool finds checksum-validated identifiers (emails, cards, IBANs...), this one finds names and real-world references.