Speech Technology


We know you want to produce the best products and for this you need to have the best technology. NowSpeak provides the very best voice interface technologies available in the market today.

We have developed our own advanced speech technologies and optimised them for embedded systems. They have been designed to be small footprint, high quality, fast and accurate.

Our speech technologies are multi-lingual and have been deployed worldwide in tens of millions of products.


Voice Trigger

Our voice trigger is designed to run continuously and listen for a trigger phrase such as "Okay NowSpeak".

We have developed unique, low power recognition technology that reliably recognises the trigger phrase, even in noisy conditions, but will not false trigger on background speech or noise.

The trigger phrase can be customised without the need for recordings.


Speech Recognition

Highly accurate recognition using our speaker independent technology which is trained to work with a wide range of accents.

It has advanced built-in noise reduction to ensure accurate recognition performance in real world conditions using single and multiple microphone configurations.

Our recognition technology is fast and responsive with recognition performed in real time. Voice commands can be changed quickly and easily.


Text-to-Speech

High quality, natural sounding text-to-speech voices for speaking text.

Our text-to-speech technology has been specifically optimised to speak people's names, device names, message sender's names and speaks phone numbers as continuous digits


Recorded Prompts

We take studio quality recordings and process them to produce low memory, high quality wideband (16KHz) and super wideband (32KHz) prompts.

With our own recording studio, equipment and access to native language experts we can get translations and produce prompts in any language you require.