13. Edgify

Founders: Nadav Tal-Israel and Ofri Ben Porat

Website: https://www.edgify.ai/

Business: Edgify use ‘edge devices’ instead of the Cloud to train deep learning models in a variety of industries.




The concept for Edgify was born through the difficult end of another chapter: Pixoneye, a Computer Vision-based product company co-founded by Ofri Ben-Porat and Nadav Israel with the ability to analyse personal photo galleries on the phone.

After three years running the business from the headquarters in London, the research team in Israel built a groundbreaking edge computing solution with far more potential than the Pixoneye product. The team later realised that they could adapt what they had developed for phones to train AI models on any hardware with a CPU (edge device), and Edgify was born in 2019.

Today, Edgify enables businesses to train their AI on the entirety of their data without the need for cloud infrastructure or big servers in store, effectively reducing the risks, costs and time associated with transferring sensitive-data to and from an external server. This technique has allowed industries like retail to train on the entirety of their data, and reach accuracy levels never achieved before, enabling self-checkout (SCO) machines to achieve 99.98% accuracy versus typical levels of 55-65% on standard machines. Furthermore, due to the distributed and continuous learning, the rates never decrease.

The legacy technology at the core of most retail stores led Edgify to identify the sector as one likely to benefit greatly from its Computer Vision solution. The product can be deployed throughout supermarkets, in any scale or scanning device – and can even distinguish between similar non-barcoded fresh groceries, such as any variety of apple. It then shares the acquired knowledge or model across a distributed yet collaborative framework of point of sale machines, leading to faster, more accurate and low-touch checkouts.



With a growing volume of distributed data-capturing devices in the market, ensuring the availability of data for real-time analysis and machine learning in a time and cost-efficient way is a mounting challenge. It is why Edgify’s edge computing framework addresses the high computation and low latency requirements for deep learning.

Edgify is currently focused on rolling out its technology worldwide across retailers’ networks of SCO machines, given the solution’s ability to achieve high levels of accuracy using any edge device with a CPU, consequently helping retailers cut item fraud at SCOs and reduce their shrink rates and prevent losses. However, the company has future plans to provide the technology for a range of industries such as health, manufacturing, intelligent homes and autonomous cars.

Edgify’s technology uses computer vision-based product recognition and is ready for full deployment across Europe to enable self-checkout kiosks, weighing stations and labelling machines throughout the store, to identify fresh produce, with zero human input on the retailer’s side.


Does Your Business Need Funding? Get Funding – Enquire Today >>>