LONGMONT — A tech startup in Longmont has emerged from stealth mode to announce its first product, a plug-and-play system that it says “drastically improves retail-loss prevention and slashes operational costs.”
DeepCam LLC, which has been in stealth mode for three years, made the product announcement Thursday at the International Security Conference & Exposition in Las Vegas.
The system uses biometrics to identify shoplifting and other suspicious behaviors. It has been successful in trials conducted with retailers in the United States and internationally.
“When you have a small population, such as an office building, facial recognition is easy,” Don Knasel, DeepCam’s founder and chief executive, said in a prepared statement. “Where biometrics fail is with large populations — tens of thousands or hundreds of thousands of people — and that’s what sets our AI technology and recommendation engine apart. Our systems have been proven to work accurately with large populations in the millions.”
The system looks for shoplifting and other suspicious behaviors that indicate further attention. A store’s staff can review these incidents, tagging those they identify as shoplifters who should not be allowed back in the store.
“Some stores may catch as few as 10 percent of shoplifters,” Knasel said. “Our system not only catches that 10 percent but is designed to drastically close the gap on what other systems miss.”
The system uses three proprietary technologies: Match — biometric identification for large populations; Index — cross-camera event association; and Advise — behavior analytics.
DeepCam began shipping its products in the fourth quarter of 2017. DeepCam’s headquarters is at 1067 S. Hover St., and it has offices in Mexico and China.