video1.0<iframe src="https://www.loom.com/embed/69b69e3801a84c5b83f72cea927252ca" frameborder="0" width="1668" height="1251" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>12511668Loomhttps://www.loom.com12511668https://cdn.loom.com/sessions/thumbnails/69b69e3801a84c5b83f72cea927252ca-adef43b2ffd328a3.gif773.845This Loom demonstrates a tool that adds an AI layer on top of digitized archives to make them dramatically more searchable and usable. It enriches handwritten notes, photos, newspaper clippings, maps, and other materials, improving transcription from about 5 percent to roughly 75 to 90 percent and generating up to 120 tags for an image in about 45 seconds. The speaker gives examples from the Pilsudski Foundation in Warsaw, focused on Polish-American diaspora and Josef Piłsudski, including semantic searches, drafting an email, and producing blog and exhibit ideas with citations to source documents. The tool can support cross-lingual queries in multiple languages and helps manage outside researchers by letting them search directly while flagging uncertain connections with confidence levels.