Deep Speech meet RASA

Voice services suite from audio streaming, hotword, ASR, NLU to ML routing and action services coordinated around an MQTT messaging protocol.

The suite is intended to support web developers using node to integrate voice services into their applications.

The suite integrates a number of services into a voice dialog flow. It uses

  • picovoice and snowboy for hotword optimised speech recognition.
  • Deep Speech for speech recognition (ASR)
  • RASA for Natural Language Understanding (NLU) and routing
  • picovoice (and soon Mimic) for text to speech (TTS)

The suite includes a React component to embed a microphone in a web page.

Thanks, but can you be a bit more descriptive of the expected interactions / collaborations between both project?

Ta, updated the original post with a bit more detail.