https://github.com/avryhof/personal-assistant

A self-hosted personal assistant framework.
https://github.com/avryhof/personal-assistant

Last synced: about 2 months ago
JSON representation

A self-hosted personal assistant framework.

• Host: GitHub
• URL: https://github.com/avryhof/personal-assistant
• Owner: avryhof
• License: mit
• Created: 2019-05-06T12:41:06.000Z (about 6 years ago)
• Default Branch: master
• Last Pushed: 2023-02-15T23:46:01.000Z (over 2 years ago)
• Last Synced: 2025-01-30T02:42:54.380Z (4 months ago)
• Language: Python
• Size: 184 KB
• Stars: 0
• Watchers: 3
• Forks: 0
• Open Issues: 5
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Personal Assistant

A simple and lightweight personal assistant.

Pretty simple -

* It listens for a wake word (just the word or "hey " or "ok ")

* Loops through skill classes until it finds a matching utterance

* Uses the Skill class parser() method to parse what you asked the skill to do

* Uses the Skill class respond() method to generate a response

* If no skills responded, it can be passed to ChatGPT and the response will be generated from there

    * There is also an NLTK responder in the works. It supports the default NLTK reflections and responders, but still

      needs a bit more development.

## Technologies

* Captures audio with the speech_recognition module

    * Currently uses the Google speech recognizer - I really want this offline, but sphinx isn't quite there, and I'm

      not really sure if I want to get into something as heavy as Vosk just yet.

* Responds with pyttsx3, so it will adapt to whatever TTS layer is on your computer

## Notes

- Should add support for [Precise](https://github.com/MycroftAI/mycroft-precise)