develops AI-based appliance that sorts email based on meaning develops AI-based appliance that sorts email based on meaning

Max Smolaks

November 14, 2019

3 Min Read

Inspired by the way the human brain understands language

by Max Smolaks 14 November 2019

Austrian startup has announced itsfirst hardware product, an enterprise appliance that can filter and routemessages by understanding the semantic content – the meaning and intent.

The platform uses unsupervised machine learning to establish what kind of information users need, and can then filter out not just traditional spam, but also marketing communications, notifications, group messages and any other email content that’s not useful to the recipient.

“The goal is to reduce the wasted efforts of handling irrelevant ormisdirected emails by first line business operations – including support,sales, purchasing,” explained Steve Levine, CMO at

The technology developed by the company differsfrom traditional natural language understanding (NLU) methods – it builds up “fingerprints”of individual words and sentences, based on their meaning, which is defined by theirconnections to other words. calls this approach “semantic supercomputing”- and the co-founder and CEO Francisco Webber told AI Business the technologyhas countless applications, from analyzing contractsand other business documents to fighting hate speech online.

Words are not numbers was founded in 2011 to commercializean approach to NLU called semantic folding, inspired by neuroscience. Unliketraditional statistical analysis NLU methods, it encodes the semantics of natural language textin a sparse distributedrepresentation called a semantic fingerprint.

“Ever-increasing unstructured data isoverwhelming the world and the available processing power and currentstatistical approaches to deal with it,” Webber said. “We are taking theconcept of supercomputing to the next level with the introduction of semantic supercomputingand the ability to deliver real-time processing of semantic content.”

The company’s Messaging ClassificationAppliance came into existence thanks to a partnership with server vendorSupermicro. The appliance is based on XilinxAlveo field programmable gate arrays (FPGAs) - accelerator cards that can bereconfigured to speed up very specific workloads.

According, the machine only needs a small number of emails to train and can easilymaster the specialized vocabulary of any businessdomain – from finance to medicine.

Once trained, the appliance can work acrossmultiple languages, including Cantonese and Mandarin Chinese. It is expected togo on sale in Q1 2020.

“Our goal is to makepossible the broad implementation and deployment of AI solutions for automatingbusiness processes and solving the most challenging use cases that dependon human understanding, decision making and execution,” Webber said.

Get the newsletter
From automation advancements to policy announcements, stay ahead of the curve with the bi-weekly AI Business newsletter.