WavyOS

Wavy

An intelligent assistant for Windows OS that combines deep automation of routine tasks, AI-powered voice control, and emotional interaction through a graphical avatar.

Our mission is to reduce the cognitive load on users when interacting with a PC and personalize the OS experience.

---

✨ Features

• Voice Activation & Control: Real-time speech-to-text recognition (STT), and text-to-speech output (TTS).
• Smart Automation & OS Interaction: Modifying audio/network settings, managing processes, emptying the Recycle Bin, and cleaning the "Downloads" folder.
• Computer Vision: Fast screen capturing, window analysis, and simulation of user actions (clicks, hotkeys).
• Avatar: An interactive overlay app avatar developed using the Tkinter library.

---

🏗️ Project Architecture & Concurrency

WavyOS uses a multithreaded and asynchronous architecture within three OS processes to keep the graphical interface smooth during heavy AI workloads.

Processes:

• Core: Listens to user commands and orchestrates other modules.
• GUI: Runs the PySide6 event loop, rendering the avatar overlay, settings window, and animations.
• Tray: Runs the pystray event loop.
• Inter-Process Communication: Implemented using ZeroMQ.

Threads:

• Worker Threads (threading & asyncio): Run the voice processing stack (RealtimeSTT / TTS) and execute Windows API or PowerShell automation scripts in the background.
• Inter-Thread Communication: Background threads push recognized commands into a thread-safe internal queue (queue.Queue), where the UI layer consumes them to instantly update the avatar's emotional state.

---

🛠️ Tech Stack

The project is developed exclusively for the Windows ecosystem.

🧑‍🍳 Multiprocessing

• ZeroMQ — a high-performance asynchronous messaging library designed for building scalable, distributed, and concurrent systems.

💻 Graphical Interface

• PySide6 — A modern graphical shell (Qt) for rendering the avatar and management windows.

🧠 Artificial Intelligence & Speech Technologies

• STT (Speech-to-Text): RealtimeSTT, faster-whisper, webrtcvad-wheels, scipy
• TTS (Text-to-Speech): realtimetts (supporting Piper, Edge, gTTS), soundfile
• LLM (Language Model): openai, pydantic
• Computations: torch, torchaudio, numpy

⚙️ OS Automation & Computer Vision

• Windows Integration: pywin32, psutil, pycaw, subprocess, Send2Trash, keyboard
• Screen Analysis & GUI: opencv-python, mss, pillow, pyscreeze, pyautogui
• Web Automation: Playwright

---

🚀 Getting Started

Prerequisites

Before installation, ensure that your Windows environment meets the following requirement:

1. Python 3.11+

Installation & Launch

1. Clone the project repository:

   git clone https://github.com/ilyamikhailov16/WavyOS.git
   cd WavyOS

2. Install all dependencies:

   pip install -r requirements.txt

3. Launch the application:

   python launcher.py

---

💻 Usage Examples

The assistant operates continuously in real-time, executing spoken commands directly.

Examples:

• «Сделай скриншот»
• «Открой Калькулятор
• «Очисти корзину»
• «Выключи звук»
• «Закрой браузер»