Home · Platform · Semantic Memory
Semantic Memory
Index multimodal timeline events and search your library by meaning — what was seen, said or read on screen — with local-first control.
Searchable timeline events
Multimodal events from Vision Intelligence, transcription, OCR and tagging are indexed in Semantic Memory. Natural-language queries surface rows of indexed files with previews and saved descriptions — models and paths stay on hardware you control.
Metadata persists in open .vtag sidecars so catalogs like NeoFinder can index results alongside originals.
Beyond filename search
Semantic Memory is the consumer-facing layer for AI-powered file search — built for archives, production libraries and research corpora that cannot rely on cloud upload queues.