What NLU and TTS/ASR engines do you use?

Natural Language Understanding

For NLU (Natural Language Understanding), we use our own solution. NLU is a subset of NLP.  It’s got a narrower focus than NLP. 

It’s based on understanding the text first (the ‘U’ of NLU), so that then the natural language processing engine (the ‘P’ of NLP) then can do its work.

We base our NLU on DeepSpeech.

DeepSpeech is an open-source speech-to-text (STT) engine from Mozilla.

It uses Artificial Intelligence/Machine Learning (AI/ML).

The AI/ML uses Google’s Tensor Flow to facilitate implementation.

TensorFlow is an end-to-end open-source platform for machine learning. It uses a comprehensive, flexible ecosystem of tools, libraries and community resources.

This allows the CX Index DevOps team to:

  • push the state-of-the-art in ML
  • build and deploy ML-powered applications.

For NLU, we use the Python Programming language. This allows CX Index to handle the data, so there's no sharing of data with any third party vendors. 

Text to Speech / Automatic Speech Recognition

For TTS (Text To Speech)/ASR (Automatic Speech Recognition), we use our own tool. So:

  • We can recognise an input's sentiment/languages.
  • Then we can extract several levels of data around it. For example, we can extract sentences/keywords/phrases.
  • Then we categorise it with predefined categories

Was this article helpful?

Can’t find what you’re looking for?

Our award-winning customer care team is here for you.