Project Goals:

  • Use fastest STT (Speech to text) / TTS (Text to Speech API) [Currently using OpenAI Deepgram]
  • Use fastest audio recording library.
  • Use a wake word, "Hey, Jarvis."
  • Response generated by OpenAI ChatGPT (tokenized to behave like Jarvis)

Virual Environment Setup For Testing

pip install virtualenv

# TO CREATE VENV
virtualenv <name>
# TO ACTVATE VENV
. .\<name>\Scripts\activate
# TO DEACTIVATE VENV
deactivate

Temporary TTS Solution

pip install pyttsx3

Current TTS Solution

Deepgram

pip install deepgram-sdk

Went down this rabbit hole to ensure things built properly.