AI-Assisted Missing Persons Web Portal

Abstract

The first few hours are crucial when a person stops appearing, and coordination tends to fail as the families, NGOs, volunteers, and police departments process disjointed reports, photos out of context, and inconsistent updates. This paper introduces an AI-assisted web portal that streamlines case intake, enforces role-based checks, and uses face-matching to flag potential leads from new sightings. It's built with humans firmly in control: AI generates ranked candidates via efficient face embeddings, but only authorized reviewers validate matches before any escalation. We cover the architecture, workflows, security features, and matching process, plus real-world tests.