Everseen Raises Over $70M for AI Tech to Detect Retail Theft

In the year In 2007, Alan O’Herley, who had previously worked with large SAP installations and retailers, was looking for a way to help retailers reduce “shrinkage,” or where a store has less inventory than it records. He settled on computer vision as a solution to the problem and founded a company called Eversen to commercialize the technology.

Everseen, which uses computer vision to, among other things, try to prevent theft of self-checkout counters – announced today that it has raised 65 million euros (~$71.32 million) in a Series A round from Crosspoint Capital Partners, a former investor in the startup. The new funds bring Ireland-based Evers’ total revenue to nearly $90 million, which O’Herlihy says is being used to grow the startup’s business through a “targeted” roadmap.

“We are experiencing strong demand for our technology from retailers struggling with the dual impact of reduced customer spending and increased operating losses,” O’Herlihy said. “The retail industry is facing challenges such as labor shortages and rising labor costs, which makes our technology even more valuable in addressing these issues.”

In particular, the downturn could hit retailers’ bottom lines as far as O’Herlihy. In the year In 2017, stores lost roughly 1.33% of revenue, totaling an estimated $47 billion, according to the National Retail Federation.

Everseen uses a combination of ceiling-mounted cameras and computer vision software to — in theory — reduce theft at brick-and-mortar stores. According to O’Herley, Everseen’s algorithms can track interest by analyzing how it interacts with items (such as SKUs) and recognizing “interest actions” by buyers and sales associates.

Beyond theft, Eversen says he can “know” when items on the shelf are out of stock and “suggest processes that require immediate attention so employees can address issues, improve trends, and reduce variances.” The platform, which processes hundreds of millions of products and tens of millions of customer interaction videos every day, can provide insights and real-time analytics from retailers’ existing tools, such as order management systems.

“All of these elements act as inputs, allowing our solution to ‘touch’ the customer to self-correct, or instruct a shop associate to engage and help,” explained O’Herley. “Our goal is to stop and recover losses, enable the retailer to intervene, promote greater customer interaction and create fluid processes, improving the overall customer experience and positively impacting the bottom line.”

Everseen has always failed in this mission. Employees at Walmart, once a major customer of Evers, told Wired in 2020 that the system often misidentifies behavior as theft and fails to stop actual instances of theft.

In response to the lawsuit, Walmart said it had made “significant improvements” to its Everseen system, resulting in fewer alerts overall. But relations between the two companies soon soured. Everseen sued Walmart, accusing the retailer of misusing the Irish company’s technology to make its own product as Everseen. (Everson and Walmart are slated for December 2021.)

Without access to the background, it is difficult to measure the validity of any system. But history has taught us that computer vision technology—especially technology designed for countermeasures—is prone to bias and other flaws.

Consider an algorithm trained to detect “suspicious” activity from a consumer. If the data set used for training is unbalanced—say, it has a large number of black shoppers stealing footage—it may point to overrepresented shoppers more often than others.

Some AI-powered anti-theft solutions, moreover, are clearly designed to distinguish a shoplifter’s gait — patterns of limb movements — from other physical characteristics. This is a problematic approach, given that disabled shoppers and others may have questionable behaviors that algorithms trained on mobile shopper videos may have.

But while we think for a moment that Everseen is largely impartial, the elephant in the room is the tracking system based on each camera: privacy. In an email exchange, Crosspoint’s Greg Clark mentioned using Everseen’s technology to capture purchasing intent and behavior “to market to certain demographics,” a tantalizing prospect, to be sure.

I asked O’Herlihy how he handles customer data, including footage that records shoppers and store associates. He said Everseen communicates with customers on its data retention policies and is – for what it’s worth – “fully compliant” with the GDPR.

Whether consumers — or affiliates — trust Everseen implicitly is another question. But potentially thorny ethical issues don’t seem to be deterring customers from signing up for the startup’s service.

O’Herlihy says Everseen counts more than half of the world’s top 15 retailers among its customers, deploying more than 6,000 retail stores and more than 80,000 checkout lines.

“As retailers look for different ways to sell and consumers look for different ways to shop, the pace of adoption of this transformative technology has accelerated,” said O’Herley. “As for technology spending, we’ve seen a shift in budgets as retailers’ challenges continue to evolve as a top priority in the industry… Everseen fits right in with current trends.

Overall, it’s true that retailers are embracing — or at least showing interest in — AI. A recent KPMG survey found that 90% of retail leaders believe their employees are ready and have the skills to adopt AI, while 53% agree that the pandemic has increased their company’s adoption rate.

In the future, Everseen — which will no doubt face pressure from rivals like AI Guardsman and VaakEye — plans to expand its technology beyond retail into sectors such as supply chain and manufacturing. The startup currently has around 1,000 employees at its headquarters in Cork, as well as centers in the US, Barcelona, ​​India and Australia, among others.

“Starting in retail, Eversen has been able to develop a foundation and library of computer vision AI for other adjacent industries,” said O’Herlihy. “Computer vision solutions are currently very siled and targeted to solve specific problems. We’re seeing a need for our platform as our customers look to solve other problems in the retail store estate.”