Deterministic Desktop Assistant

The assistant can now read and execute from an external function library when prompted by user voice input. The assistant can handle multiple closely named functions or input that includes multiple function names by providing an options list to the user, allowing them to choose through their voice. The functions used in the video are placeholders/demos because eventually the user should be able to write their own scripts that suit their own needs or find community scripts online to import. My next step is probably to try making an app to house the program and give the assistant a graphical interface for configuring settings, writing/importing scripts, and other customizations.

A decent amount of the time I spent working on this project so far comes from learning how to use Git with my code. Between this devlog and the last, I redid the entire Github setup for this project. This has been my first time ever using Github from the IDE so the learning curve was initially a little steep, but I think I’ve got the hang of it now!

I finally picked up a project I had started about a year ago where I decided to make a personal virtual assistant that didn’t need to use AI to do anything. I’m still deciding on a working name for the assistant, but the project now has basic voice-to-text functionality, a voice playback function, and all code-breaking bugs seem to have been fixed. My next step will be to add a few testing functions and to find a way for the program to decide which functions I want to run based on the user’s voice command and the given names of the functions. The assistant currently only runs through the terminal, but eventually I want to be able to manage the assistant through a GUI.