Overview
- The platform is already in daily internal use at the Academy and is structured with a private workspace and a forthcoming free public portal on the RAE website.
- Core modules include a Word Observatory, automated linguistic queries, verification tools that suggest corrections, a diversity collector, and the digitization of historic “red-dot” lexicographic cards.
- The Word Observatory is described as capable of scanning roughly one million expressions per day across media and online sources to surface neologisms, regionalisms, and terms outside the dictionary.
- RAE emphasizes human oversight: the system classifies and proposes answers but routes unresolved questions to linguists, and it does not auto-correct user text.
- The effort is backed by agreements with Google, Microsoft, Telefónica, and AWS, about €5 million in investment, and the ALIA family of in-house AI models developed using a top-tier supercomputer.