While image recognition in photos and video has become increasingly sophisticated at an incredible rate, Standard Cognition co-founder Jordan Fisher says everyone is chasing autonomous driving — or, the "shiny object," as he calls it.
That's not the aim of Fisher and his team, who set out to start a company that's focused on streamlining the checkout experience. That's the end result of Standard Cognition, a startup that provides technology that helps businesses see what people are picking up off the shelves — and eventually purchasing. The end goal is for someone to walk in, grab a jar of mustard off a shelf, and walk out without having to wait in line to pay for it. And the idea of just turning every store into a pantry is such low-hanging fruit that the company has been able to attract new financing.
"The research is so amazing, the progress has been so incredible — but what we really saw was despite all this tremendous progress and adoption of all the major tech companies in the world, it hasn't transformed the world yet," Fisher said. "It makes Google search better, whatever, fine. But machine vision is the key to unlock the digital world from the physical world. We wanted to find what that angle was, and where we could apply this technology [in a way that it would] have a huge impact. We were looking for something that was challenging enough so we had a moat for defensibility."
Standard Cognition said today it has raised another $5 million in a round led by Charles River as well as Initialized Capital, Y Combinator, and other smaller investors. CRV's Devdutt Yellurkar led the firm's deal in Standard Cognition, with Garry Tan working with the founders as well.
The goal of Standard Cognition is to plug into some slightly more sophisticated cameras than the average security camera right now, and then basically just give the average human an app and say go. The cameras identify who walks in, what they're carrying, and what they end up walking out with and bill them for those products. There's no facial recognition that goes into it, working in such a way that tries to keep the data anonymized and then removed once the person leaves the store. Part of the way it looks to do that is by hosting all of its processes on-premise, rather than over a cloud-based infrastructure.
"There are always bigger competitors, more well-financed companies," Yellurkar said. "I think these guys, like every good startup, are singularly focused on solving one specific problem. They are solving for small-format retail and trying to get a zero wait-time checkout experience. That focus is going to give them the opportunity to run faster. They're not solving tech for the sake of tech. They're using tech to solve a specific problem."
Technology like Standard Cognition's seemingly has obvious applications for larger retailers, which Fisher says would "love to" build in-house but have trouble competing for talent and getting it off the ground. But another natural application is smaller retailers, or even the average mom-and-pop or corner store, which can get people in and out faster and help drive more business to those locations based on just the convenience. There's a bit of a barrier to entry on the user behavior side with downloading the app, but after a few tries people get used to it pretty quickly, he said.
If this sounds familiar, it probably should — and in your recent memory too. There's a startup called Bodega that sparked some controversy that was looking to build a process somewhat similar, though it was more geared around having a distributed cabinet network that people could just walk up to and snag said jar of mustard. Advances in machine vision have enabled a whole class of startups that can use the technology to pick off the kinds of low-hanging fruit that Standard Cognition wants to grab, which could lead to some significant competition for the company going forward.
"It's just gonna be a huge Cambrian explosion that we're starting to see," Fisher said. "I think it's cool, there are things like Focal Systems [or Bodega] that are working on other ways to do check-out. All those are probably gonna have some part to play in the future of retail. But I also think retail is so massive and the existing infrastructure of retail, the 10,000 store fleet of 7 Eleven, or other stores even in Asia. If you want to reach that part of retail, and if you want to repurpose the existing infrastructure, I think you have to have something like this."