Early stage: Software that can tell a friendly neighbor from a criminal

Who they are: Deep Sentinel, a Pleasanton-based home security startup backed by Shasta Ventures.

What they do: Use artificial intelligence to patrol the outside of your home and alert you to potential intruders.

Why it’s cool: Deep Sentinel pairs AI with off-the-shelf cameras. The startup’s software recognizes potential threats in the video footage — distinguishing someone stealing your mail from a neighbor walking their dog, for example — and alerts you to them in real time. That means unlike with a traditional security camera, you don’t have to spend all day watching the video footage for it to do any good.

“Most security cameras today are what we call ‘crime recording devices,’” co-founder and CEO Dave Selinger said, “and crime recording just isn’t very valuable.”

Deep Sentinel’s technology also is designed to eliminate the false alarms that plague other home security systems. In 2015, the Santa Clara Police Department received 3,591 dispatch requests from triggered burglar alarms. Of those, only 35 were valid — the rest were false alarms, the police department wrote in a post on neighborhood social networking site Nextdoor.

Deep Sentinel’s technology is about 99 percent accurate, Selinger said. It uses the same type of technology that self-driving cars use to navigate objects in the road, and Facebook uses to identify people in photos.

Ultimately, Deep Sentinel hopes to not only alert you to threats but also to issue some kind of deterrent to scare away bad actors. That could be anything from turning on automatic sprinklers to sending in a remotely piloted drone.

Where they stand: Deep Sentinel raised $7.4 million in a series A led by Shasta Ventures in April. The company plans to launch a new version of its product for beta testing in three to six months, and to bring it to market early next year.

Read more at Mercury News

Trackback from your site.

Leave a comment

Save my name, email, and website in this browser for the next time I comment.
Share via