Deeply unsettling startup installs cameras to scrutinize drive-thru workers

Restaurant tech startup Agot AI uses “computer vision” to spy on employees.

We may earn a commission from links on this page.
Person receiving food at McDonald's drive-thru
Would you like a side of dystopian monitoring with that?
Photo: Ezra Acayan (Getty Images)

Last weekend, my boyfriend and I cruised through the Dunkin’ drive-thru. I ordered an everything bagel with cream cheese, and he ordered a bacon, egg, and cheese breakfast sandwich. Upon receiving our order, we realized that the drive-thru attendant had accidentally combined the orders, giving us a bacon, egg, and cheese breakfast sandwich on an everything bagel with cream cheese.

No biggie. I popped inside, explained the mistake, and they fixed our order, leaving us with an extra breakfast sandwich. Took all of three minutes. But now, a sinister tech start-up is dedicating millions of dollars to “improving fast-food order accuracy.” Why? Because they can, I guess.

CNBC reports that restaurant tech start-up Agot AI has closed a $10 million funding round that will support the company’s mission: to scrutinize workers and determine whether they are preparing orders correctly. How does the company do that? Through the most invasive means necessary, of course! Agot AI installs overhead cameras in restaurant kitchens and uses “computer vision” to spy on workers, presumably to improve labor efficiency. Agot co-founder and CEO Evan DeSantola told CNBC that the technology can spot over 85% of order errors, calling attention to the errors before workers actually serve the food. I don’t know exactly how that works, but I imagine it’s through some sort of horrifying, demoralizing alarm system.

Advertisement

“We see that across the [quick-service restaurant] industry order accuracy is becoming an increasingly large problem as a result of the shift to drive-thru,” DeSantola said. “What was once a smaller pain point, when the accuracy rates haven’t gotten much better, is now a much larger pain point.”

Is it? Is it that much larger of a pain point? To be mildly inconvenienced by a very understandable mistake made by an exhausted fast-food employee working long, hot hours for a pittance? I don’t think so, but DeSantola does say that Agot’s typical customer has at least 2,000 restaurant locations. (He declined to share the names of current restaurant clients.) “What was very clear to us about Agot is that it is not a point solution,” said Greg Golkin, managing partner of equity firm the Kitchen Fund. “It is a platform that is being built, and order accuracy is just the first application. Computer vision is not going to stop there.” Well, uh, have a great weekend, everybody. There’s some dark stuff on the horizon.