Innovation intelligence brief
Innovation intelligence brief
A data analytics SaaS company, provides computer-vision technology that can analyze a customer's public online photo galleries to create a personal risk profile.
What became of it
Completed · Medium confidence · Last verified 2026-10-17
As catalogued
London/Israel computer-vision SDK that scanned a user's on-device photo gallery and extracted abstract signals (life events, interests) to build personal risk and marketing profiles for insurers, banks and advertisers, without uploading the photos.
2026 reality
The Pixoneye team now operates Edgify, a London-based edge-AI company co-founded by Ofri Ben Porat and Nadav Israel (CEO in 2026) that trains vision models on in-store devices, mainly for grocery self-checkout loss prevention. In August 2026 Edgify raised a USD 9M Series A+ led by Rank Ventures and Mangrove Capital (total funding about USD 25M) and reported deployments in 2,042 stores. The pixoneye.com domain is now an unrelated gadget-review site.
Pixoneye, which never found a durable market for photo-gallery profiling, repurposed its on-device technology for retail and launched as Edgify in 2019; it raised a USD 6.5M seed in October 2020. The insurance profiling product is no longer offered.
The on-device processing know-how from the privacy-sensitive gallery-scanning SDK was redeployed to a broader edge-AI training market; privacy constraints on personal image data plausibly limited the original use case, but no company statement on the motive was found.
Historically notable as an aggressive use of personal image data for insurance and marketing profiling; the successor business has no insurance focus.
Third taxonomy
REVIEW · Verified the Pixoneye-to-Edgify link and August 2026 USD 9M round via The Next Web and Tech.eu. Corrected the CEO (Nadav Israel, not Ofri Ben Porat), store count (2,042), pivot dating (Edgify launched 2019) and a source title that claimed 'formerly Pixoneye' though the page does not say so; added Start-Up Nation Central.
Documented evidence
Homepage
Homepage
Insurance
Research interpretation
Classification paths, as catalogued, connect this example to the original insurance digitalization research models.