Corporate filings rarely tell dramatic stories, but the paperwork can reveal where specific companies are finding capital—and Edgify, a London startup in retail technology, has been raising funds.
London's Edgify, a company most shoppers have never heard of but whose software may be watching them scan groceries, has quietly raised fresh capital. The evidence: a trail of share allotment documents filed with UK regulators on February 24, March 11, April 2, and May 6, 2025, followed by a September 22 filing naming Mangrove V Investments S.A.R.L. as a significant stakeholder. How much? The company isn't saying. Valuation? Also undisclosed.
What Edgify will say is that its edge-first AI platform now runs across more than 8,000 devices in a dozen countries, scanning for theft at self-checkout lanes without ever touching the cloud. That footprint matters in an industry wrestling with what some estimates peg as a $142 billion annual shrinkage problem worldwide—though that figure, like many in retail loss prevention, carries its share of caveats and assumptions.
The pitch: Keep the AI local
Edgify's argument is simple, almost elegant. Why send checkout video to the cloud when you can run computer vision right there on the scanner?
The company retrofits existing hardware—standard USB cameras, scales, barcode readers—with software that watches for barcode switching, phantom scans, and the mismatch between what the camera sees and what the register records. It calls the architecture a "distributive collaborative framework," which in plainer terms means the system learns on-device, across thousands of locations, without pooling sensitive video in a central data lake.
The rationale hinges on latency, bandwidth costs, and a growing unease about surveillance creep. For retailers already nervous about customer backlash—self-checkout is hardly beloved—keeping footage local has appeal.
Company data accessed in August 2026 claims the platform has processed north of 400 million samples across 2,042 stores, with precision hitting 98.58%. Those numbers are company-reported and undated, so the usual grains of salt apply. Still, partners have vouched for the technology in public. Dell Technologies published a validated blueprint for deploying Edgify on its NativeEdge platform. Elo Touch Solutions wrote approvingly in a May 2026 blog post about preserving the self-checkout experience while cutting losses.
Follow the filings
The new funding builds on a $6.5 million seed round announced back in October 2020, led by Octopus Ventures and Mangrove Capital Partners. Also in that round: an unnamed semiconductor player and HBS Alumni Angels. Mouro Capital, the fintech investor, lists Edgify in its portfolio too, though it hasn't disclosed terms or timing.
UK Companies House records show Mangrove V Investments formally registered as a person with significant control on September 12, 2025—reinforcing continued involvement from a firm that's been there since the beginning. Octopus Ventures still lists Edgify among its active bets.
The company was co-founded in 2014 by Ofri Ben-Porat and Nadav Israel. Ben-Porat's directorship ended in April 2024, according to filings, though no public explanation has surfaced. Current directors include Nadav Tal Israel, Mitchell Bayer-Goldman, Hans-Jurgen Schmitz, and Gerard Lopez.
The self-checkout paradox
Retailers love self-checkout for the labor savings. Shoppers tolerate it—sometimes grudgingly. And thieves, apparently, have taken note.
A June 2026 study from the University of Leicester linked rising losses to the spread of automated lanes, adding academic weight to what industry insiders have muttered for years. Hardware partners like Zebra and Elo now routinely emphasize in-lane computer vision as a counter, and the market for such solutions has gotten crowded.
Edgify competes with Everseen, which published a widely circulated 2025 report on retail AI adoption, and Standard AI, which leans on Google Cloud for its vision applications. Edgify's edge-only stance positions it as the privacy-conscious alternative—particularly for chains wary of real-time cloud inference or jurisdictions with strict data residency rules.
The company demonstrated its technology at NRF 2020 as part of an EdgeX Foundry consortium that included Shekel, Flooid, HP, IOTech, and Intel. It returned to NRF in 2026, this time exhibiting alongside Zebra and Elo, and made an appearance at the Retail Technology Show last April.
What comes next?
Details on how Edgify intends to deploy the recent capital remain elusive. The company's federated learning model—training algorithms collaboratively across thousands of devices without centralizing data—suggests room for geographic expansion, but public statements on growth strategy or customer pipelines have been thin.
With retail AI spending climbing and self-checkout technology under pressure to deliver both efficiency and security, Edgify faces a familiar challenge: turning device count into durable market share. Deployment numbers look good on paper. Revenue, customer retention, and competitive moats—those are harder to reverse-engineer from corporate filings.
Perhaps the company prefers it that way. In an industry where the best technology often goes unnoticed by the end user, flying under the radar might just be the strategy.
