Skip to content

Refract Speech: The Sentient Search Engine

Ever wish your search engine knew what you were talking about? Intelligent Search engine Refract Speech ‘reads’ webpages

Being slightly dyslexic, CEO of Refract Speech Benjamin Felsham always felt search engines fell short of his expectations– they never quite hit the nail on the head unless you know just how to use them.

So for the past 14 years he’s been developing his own search engine.

Refract Speech, an anagram of ‘perfect search’, uses artificial intelligence to ‘read’ webpages and actually learns about topics, according to Felsham, who is trying to raise £210,000 through Kickstarter in London to fund the project.

Watch their campaign video:

Sentient Assistant

Benjamin Felsham compares the way his system works to having a knowledgeable assistant in a library– something that knows what’s in the books and can point you in the right direction.

As the system learns, you are searching its knowledge rather than the webpages themselves, which makes it a fundamentally different kind of search, according to Felsham.

However entering the search game will inevitably pit Benjamin Felsham against tech giants Google and Apple, who each have impressive offerings in terms of personal assistants in the form of Google Now and Siri.

If Refract Speech works like Felsham claims, it would realise what the Semantic Web or ‘Web 3.0’ failed to deliver what was promised years ago– a search that understands what you are asking.

But with the progress of  speech recognition and the ability of personal assistants to harness all sorts of data to bring you highly relevant information, Refract Speech will need to do something truly amazing to be able to compete on any kind of level.

If you think search could be better, and want to give a leg-up to the underdog, then check out the Kickstarter page and get pledging, as £210,000 is not going to raise itself.

 

 

Topics

Register for Free

Get daily updates and enjoy an ad-reduced experience.

Already have an account? Log in