WHO WE ARE

Talking 2 Rabbit Sarl is a startup company, founded in Luxembourg, operating in the domain of Voice Recognition Technology. Our vision is to create the most innovative algorithms for Voice Recognition industry that pave the way for the new telecommunication services and the wide-spread use of voice-enabled devices.

We have a number of mathematical geniuses on board  who are working hard to develop and offer the beyond state-of-the-art innovative algorithms. Those algorithms are the basis for the next generation software applications and devices to make human life easier. Our revolutionary method will offer a dramatically enhanced quality in the case of cross-language voice commands – i.e., Italian speaker pronounces a British location or family names.

Our byproduct of the new methodology is lowering of the data amounts requirements with all beneficial consequences (less energy consumption, cost and time).



CONTACT

1, Cote d'Eich
L-1450 Luxembourg
Luxembourg
Tel: +352 691 454 014
hello@talking2rabbit.com

Good friends, Martin (passionate PhD researcher in speech enhancement) and Michael (project manager
with a solid background in business development & marketing) met together to talk about
one innovation that Martin had been working on for a long time.
His idea was simple… to create an algorithm that would be a game changer.
Human foreign accents in any language are a serious barrier to growth of voice technology.
Accents prevent millions of users from communicating effectively with software.

In 2016 we got to work. We started by approaching the inaccuracy in recognizing
geolocations and family names spoken by non-native speakers.
Accents were severely downgrading the results. Through the use of advanced mathematics,
Markov's chain algorithms, and several high-end servicers, we brought together American native speakers
and had them read and record German geolocations and German family names.
This work resulted in the creation of two algorithms: “ZOOMER +” and “RABBIT EAR”.
These algorithms treat accents and speech irregularities similar to environmental noise.
While traditional algorithms don't recognize a voice command and return any response they can think of,
these new algorithms don't give up so easily, searching deeper into an extended tolerance belt.
And we did this by using the right technology and the right people in the business.

We continued toward our goal throughout 2017.
Each day we painstakingly planned and tested different scientific hypothesis
while adhering to ITU-p835 standard regulations.
  And all our hard work couldn't have been at a better time.
This was the year “the Economist” chose voice recognition technology as the technology of the year.

Then, finally, in 2018 we received confirmation from the European Patent Office in Munich, Germany,
that our invention is unique.
Our patent, after long last, was pending.
Now it's time to show our invention to the world.

Our enthusiasm while creating and testing our further algorithms is still very high.
Martin our CTO has another great idea also.
We are in advanced talks with a global producer of hearing aids to implement
our ZOOMER + in them to reduce outside noise for their devices.
The scientific work goes on.