An empty space on a supermarket shelf is a small mystery with an impatient audience. A customer wants an item. A store may believe it has that item. Somewhere between the inventory record and the aisle, the two versions of reality have parted company. Luis Vera has spent much of his working life trying to arrange a reunion.
His pursuit began in 1992, while he was still at university. It has involved cameras mounted above supermarkets, fixed cameras studying shelves, a company that closed, and eventually a robot that travels through a store taking pictures. The machinery changed. The awkward gap between what a retailer knows and what a customer can buy kept providing him with work.
Vera is a Chilean industrial engineer and co-founder of Zippedi, alongside Ariel Schilkrut and Álvaro Soto. He was its CEO when it raised its Series A in 2022, and the company continues to list him in that role. Yet the financing round is a late chapter in a story whose early chapters contain rather less applause.
The camera came before the robot
Vera grew up in Silicon Valley, returned to Chile to finish secondary school, and studied industrial engineering at the University of Chile. His career would keep connecting those two places: Chile as a place to build and test retail technology, California as a place to pursue customers and capital.
His first company, Prospect, began with closed-circuit television in supermarkets. Cameras could help retailers investigate losses and understand internal processes. This was a practical starting point for a young entrepreneur. Stores had valuable goods, complicated operations, and plenty of things their managers could not watch at once.
By 2006, Vera and Schilkrut were building SCOPIX. Fixed cameras offered another way to turn the physical shop into information. Their ambitions extended beyond surveillance to the everyday workings of retail, including queues and product availability. A video feed could contain clues about how a store was functioning, provided someone could extract them economically.
The qualification mattered. Images are easy to collect compared with the work of interpreting them across a large chain. Early computing and camera systems imposed limits on the business the founders wanted to build. The idea had commercial appeal; making it repeatable at a workable cost proved much harder.
A rejection with a return ticket
One of Vera’s consequential introductions happened well before Zippedi existed. In 1998, Jason Green and Endeavor co-founder Linda Rottenberg interviewed him at the organisation’s first international selection panel. He did not pass. The panel nevertheless saw enough promise to create a comeback opportunity, giving him support to address weaknesses in the business model.
Vera returned to a panel in Argentina in 1999 and was selected. Years later, Green would invest personally in Zippedi. It is an unusually long arc for a professional connection: an evaluator who first declined to select him eventually backed a later company.
The episode gives perseverance a more useful meaning than simply refusing to leave the room. Vera went away, accepted advice, and returned with work done. The relationship survived an unsuccessful first attempt. A network built around that possibility can matter as much as an introduction made at precisely the right moment.
When the shelf finally started talking
SCOPIX eventually closed. Vera continued pursuing retail technology in California, including virtual-reality approaches to shopping, while Schilkrut worked in Chile. Their contact continued. At the University Católica, Soto and his research team were developing the navigation and visual-recognition technology that would give their next attempt a different foundation.
A commercial opening with Jumbo brought the strands together in 2017. The supermarket wanted a proposal involving a robot. Schilkrut visited Soto’s laboratory, where work on recognising supermarket shelves was already under way. Rather than begin every component from scratch, the founders could combine retail relationships, operating experience, and university research.
The initial paid pilot concentrated on shelf-price checks. Its scope was concrete: compare what the shelf displayed with what the store’s system said. A discrepancy could become a task for somebody to fix. After years of trying to capture the shop, there was a bounded problem, a customer willing to test a solution, and a machine capable of making the rounds.
That combination explains why Zippedi’s origin belongs to three founders. Vera brought a long acquaintance with retailers’ problems. Schilkrut brought his own experience developing and delivering solutions. Soto’s university work supplied the robotics and artificial-intelligence research. The company emerged from their ability to put those pieces to work together.
Tito takes the night shift
At Sodimac, employees gave the robot a name: Tito. There is something pleasingly sensible about this. A machine that keeps turning up for the night shift is easier to discuss when it has a name shorter than its technical specification.
Sodimac began with a pilot and worked with Zippedi to improve the service before announcing a rollout to all its Homecenter and Constructor stores in Chile in December 2021. The robot checked displayed prices, missing products, and items in the wrong places. Staff could start the day with information gathered while the shop was quiet.
The relationship also placed a limit on what automation meant. The robot collected observations. People still needed to correct prices, replenish shelves, and deal with customers. A useful scan had to end in work that somebody could understand and carry out. Otherwise, the store would simply possess a more elaborate account of its own mistakes.
Zippedi’s university roots received recognition in July 2021, when the company won a Chilean technology-transfer award for a scientific and technological venture. The award acknowledged the difficult journey from laboratory results to a product operating in shops. A functioning demonstration and a dependable service are separated by many unglamorous details.
“El fracaso es una parte importante de la ecuación del éxito.”
Luis Vera, 2018. “Failure is an important part of the equation of success.”
A digital twin earns its keep
Vera’s larger ambition is a digital twin of the physical store: a structured representation of products and their positions, assembled from observations on the shop floor. Price checking is one use. Stock availability, assortment decisions, and finding products are others. The same underlying view can serve several people with different jobs.
This helps explain the commercial model. Zippedi offers robots as a service, with applications built around the information they gather. Retailers and suppliers both have reasons to care about a shelf. The retailer wants a well-run store; the supplier wants to know whether its product is present and properly displayed.
For someone ordering groceries remotely, the problem becomes particularly plain. The selection shown on a screen may differ from the selection available in the shop. A substitute item or refund is the consumer’s experience of an information problem deeper inside the operation. Vera’s work approaches that problem from the aisle outward.
By 2022, the engineering behind the robot also reflected an old lesson about computing capacity. Zippedi used Google Cloud to process shelf scans, with machine-learning models identifying products and outdated labels. Moving infrastructure into the cloud reduced the burden of managing another layer of physical equipment. The founders already had robots to build and maintain. Every additional system competed for time and money that could go into improving the service. For a business shaped by earlier technical limits, that allocation mattered.
The company’s current platform describes that sequence as capture, process, and act. Its Bruno app turns shelf information into prioritised tasks for store associates. For Vera’s long-running project, the significance is straightforward: the image has acquired a route to the person who can do something about it.
The distance from Chile to Japan
In September 2021, Zippedi announced a $6.9 million seed round. A $12.5 million Series A followed in April 2022, led by Transpose Platform. Vera’s decades of work now sat behind a company receiving capital to expand its approach to store digitisation.
Expansion brought new tests of the economics. In April 2024, Zippedi described an agreement to put robots into six Unimarc supermarkets that year. Vera explained that smaller stores made financing the technology more difficult. The company developed a platform for that segment that allowed interaction with suppliers, including information and task generation.
Japan supplied a more personal moment. After closing a contract with a Japanese home-improvement chain, Vera and Schilkrut got into a car. Schilkrut asked whether Vera had ever imagined they would export robots from Chile to Japan. Vera later recalled the exchange with evident pleasure. A career spent crossing between Chile and California had acquired another destination.
On October 2, 2026, Vera appeared on Radio Duna’s entrepreneurship programme, introduced as Zippedi’s CEO and an Endeavor entrepreneur. The description connected two strands that had been running together for decades: building a company and learning how to return after an early disappointment.
His aspiration remains rooted in the shop itself. He wants retailers to operate with digital twins that help them see and correct the details of daily trade. It is a patient ambition for a field fond of sudden arrivals. The customer at the end of the story has a considerably shorter horizon: find the right item, at the right price, and get home. Vera’s thirty-year effort meets that customer in the aisle.
More from the aisle
Conversations, early chapters, and the work today.