R Majestic Studio

ENKI Second Memory explores a user-controlled visual memory for everyday life: “Where did I last see my keys?” The goal is to connect what a person authorizes their Ray-Ban Meta glasses to observe with a searchable, evidence-based memory. ENKI’s existing private iOS application provides the camera integration, authentication, storage infrastructure and Spanish voice interface. During the hackathon, we developed the experimental Second Memory observation module, separate consent for remote analysis, the secure Meta Model API adapter/backend, structured visual observations, explicit local saving and deletion, and a generic Voice → Second Memory query bridge. The independent public repository contains the reproducible core and tests; the README distinguishes prior infrastructure from new work and explains its scope. The design is Ray-Ban Meta → ENKI iOS → authenticated HTTPS backend → Meta Muse Spark 1.3 → validated structured scene/object data → optional user-authorized local memory. A query searches saved observations and should return the latest relevant evidence, time and context, rather than inventing a location. API credentials stay on the server. Observation is visible, time-limited and stops under privacy/lifecycle conditions; images are not continuously sent to Meta. Physical glasses-to-iPhone frame reception and sampling have been demonstrated. A genuine Muse Spark 1.3 test inference has also been verified separately. The software components and automated tests exist, but the complete physical capture → analysis → save → spoken retrieval workflow is still being integrated and has not been validated end to end. The video presents this prototype status honestly, in Spanish with English subtitles. Our vision is a natural combination of vision, memory and voice that helps with everyday forgetfulness while keeping people in control of what becomes a memory.