This research aims to develop a missing person search system based on facial recognition technology to enhance the effectiveness and efficiency of identification. The system utilizes FaceNet to extract unique facial features from uploaded images, supported by OpenCV (haarcascade_frontalface_default.xml) for initial filtering, ensuring only images with detected faces are processed into the database. For managing large datasets, the HNSW (Hierarchical Navigable Small World) algorithm is implemented for fast indexing, while FAISS (Facebook AI Similarity Search) accelerates feature matching within extensive datasets. The system is designed as a Progressive Web App (PWA) with a user-friendly interface, accessible across various devices. Testing was conducted at the Kudus Police Department, yielding high identification accuracy and significantly faster search times compared to conventional methods. The PWA implementation ensures flexibility and ease of user access. This study concludes that the integration of modern technologies such as FaceNet, HNSW, and FAISS effectively supports missing person searches. These findings contribute significantly to the development of technology-based solutions for handling missing person cases.