Press "Enter" to skip to content

Edgify: London AI startup transforming collaboration on edge devices

• Edgify is pioneering an alternative method of training deep-learning and machine-learning models at the edge.
• The funding will help Edgify expand into new verticals and new geographies beyond its current focus on retail.
• “We had the support of existing investors for this round so spent our time this year searching for our strategic investor,” Ofri Ben-Porat, CEO and cofounder of Edgify, told Business Insider in an interview. “It took a good six months of negotiation but their roadmap is aligned with ours and we can expand our verticals now.”
London-based AI startup Edgify has raised a $6.5 million seed round from an unnamed strategic investor (referred to as a semiconductor giant) alongside Octopus Ventures and Mangrove Capital Partners.
Edgify was previously known as Pixoneye, and pivoted from a previous data analytics focus in 2018.
The startup is now pioneering the training of deep-learning and machine-learning models on edge devices.
An edge device can be anything from a phone to a smart fridge and Edgify says its software can effectively teach devices what to look for and then collaborate without sending information to the cloud.
Edgify says the benefit is a reduction in the risks, costs, and time associated with transferring sensitive data to or from an external server.
For now the startup is focused on retail, but wants to expand to the medical and automotive industries with this funding.
For example, the firm has a partnership with point-of-sale maker Shekel, embedding grocery scanner tech that can recognize different types of fresh produce on checkout.
“We have a two-layer software platform which sits on each point of sale at a retailer and reads the detail on each inputted data point before allowing those various checkouts to talk to each other and teach each other the model,” Ofri Ben-Porat, CEO and cofounder of Edgify, told Business Insider in an interview.
Currently, item recognition at a checkout for a piece of fruit or vegetable, for example, is between 65-70% whereas Edgify’s solution increases the accuracy rate to 99.98%, Ben-Porat said.
It could potentially be used to reduce fraud in retail and would theoretically prevent cases like the French teenager who bought a Playstation 4 for the price of nine apples by fooling the weight scanner.
“We had the support of existing investors for this round so spent our time this year searching for our strategic investor,” Ben-Porat said. “It took a good six months of negotiation but their roadmap is aligned with ours and we can expand our verticals now.”
The company will expand further into the US, Europe, and Israel as well. Read from source….