Skip to content

General purpose assistant powered by OpenAI API

License

Notifications You must be signed in to change notification settings

Aktyn/aktyn-assistant

Repository files navigation

Aktyn Assistant

Current Version GitHub license GitHub issues GitHub Stars


Overview

Aktyn Assistant is an application that allows you to interact with an AI on various types of devices while performing regular tasks.
It can be easily activated by a keyboard shortcut.

(Coming soon) It is able to take a quick glance at your screen and answer questions about it.
It can be easily activated by custom shortcut or voice command.
By utilizing different type of AI models, it can perform different tasks like generating images, making real time conversations, understanding image context, etc.

Requirements

Upon first run, you will be prompted to enter your OpenAI API key.
If you already have an OpenAI account, you can generate an API key here.


Development setup

Prerequisites

  • yarn install - yarn 4.2.2 or newer is recommended

Terminal app

  • yarn build:all and yarn start:terminal to run the application with terminal interface
  • Some console features doesn't work inside turbo which handles the development run.
    To make sure the console features work while you develop terminal app you can run yarn dev:packages to watch changes only in packages/ and then yarn run build && npx cross-env NODE_ENV=dev yarn start from apps/terminal directory

Desktop app

  • yarn build:all and yarn start:desktop to run the application with desktop interface
  • yarn dev:packages to watch changes only in packages/ and then yarn dev:desktop
Building desktop app for various platforms
  • yarn build:all and yarn start:desktop will build project binaries and prepare them for distribution (check apps/desktop/out directory afterwards)
Publishing desktop app
  • yarn build:all and yarn publish:desktop to build and publish the application to github releases

Future plans

  • Speech synthesis and recognition
  • Attaching screenshot or selected screen region to active chat with AI
  • Real time voice chat utilizing GPT-4o model possibilities
  • Support for multiple AI providers
  • More integration with system (eg. ability to analyze playing audio and answer questions based on it)