The client - a police agency in the Arabian Peninsula - was struggling to efficiently and effectively process the over 10,000 reports of lost/found items being filed monthly by the public. High levels of human error in documenting item descriptions contributed to bottlenecks and the unsuccessful matching of lost reports and found items, impacting resource allocation and public trust. 

As a proof of concept, Siren built a system to streamline lost and found operations. It included the following technologies:

• AI image captioning that automatically generates accurate, detailed descriptions of items using visual recognition, thereby reducing human error. 
  

• An AI inventory classification system that automates and streamlines database management, enhancing searchability and matching precision. Implementing AI in inventory management practices like item-level tagging can increase inventory accuracy to 95%. 
  

• AI similarity matching that compares lost and found items, thereby accelerating item reporting and retrieval for the public.
  

• An interactive chatbot to guide users through the reporting process and answer any questions. 
  

The tool was designed to integrate seamlessly with the client’s existing on-premise systems and applications. 

In benchmark testing against manual processes, Siren found that the system reduced the time spent matching found items with lost reports by 95%.

High resolution image matching