Startup Vaak offers AI surveillance software to predict shoplifting before it happens

Japanese startup Vaak has developed artificial intelligence software it says can analyze behavior in footage from retail surveillance cameras to predict shoplifting, Bloomberg reports.

VAAKEYE was launched to beta in March 2018, and has been tested in stores for several major retailers. The company says the trials have shown a 77 percent reduction in shoplifting, and a suspect was arrested with the technology’s help after allegedly shoplifting from a convenience store last year. Tyco Retail Solutions’ research indicates that shoplifting amounted to $34 billion in lost sales in 2017, or roughly 2 percent of global retail revenue.

The company is currently testing in a few dozen stores in the Tokyo area, and Founder Ryo Tanaka told Bloomberg it is hoping to launch its production software this month, and be deployed to 100,000 stores in Japan within three years. VAAK is currently seeking to raise 1 billion yen (US$9 million) in Series A funding, having already raised 50 million yen ($450,000) from SoftBank Group’s AI fund.

VAAK says in a company announcement translated by Google that it is seeking sales partners, and will target convenience and drug stores, along with the broader retail market. Eventually, the company hopes to apply its technology to other social issues, such as public safety or suicide prevention.

FaceFirst CEO Peter Trepp told Biometric Update last year that most retail deployments of facial recognition in the U.S. are for loss prevention, but that other applications, such as customer loyalty programs, could soon follow.

Article Topics

artificial intelligence  |  criminal ID  |  Japan  |  retail biometrics  |  surveillance