Abstract
Urarina is an endangered language spoken by a small indigenous community in Northern Peru. In 1987, a research team from the University of San Marcos in Lima, Peru, conducted a documentation study to help preserve the language, producing thousands of handwritten lexicon cards and cassette recordings of native speakers. Decades later, these materials had deteriorated significantly, making them difficult to use in contemporary linguistic or educational research. Therefore, this project aims to digitize and restore these artifacts to inform both modern linguistic analysis and future educational initiatives. By converting handwritten notes into a structured database and enhancing analog audio recordings, the project provides invaluable data to support national educational program proposals and to lay the groundwork for integrating cultural knowledge with emerging AI technologies. Beyond safeguarding the materials themselves, this work highlights the role of engineering, particularly AI, in revitalizing pre-digital studies and expanding around endangered-language preservation. The methodology and results presented here aim to encourage similar restoration efforts and to contribute to the conservation and potential pedagogical use of endangered languages.