Jazmin — A Digital Personality Assistant

Jazmin is a Windows desktop emotional assistant I built in Python with a custom GUI, real-time TTS, and strict audio-playback control (no overlapping lines).
Download the app: https://jazminpy.com
This repo showcases the source.

Features

• Custom Tkinter GUI with image-based controls
• OpenAI-powered dialogue (configurable)
• Voicemaker TTS with exclusive playback (queued / overlap-free)
• Idle timeouts, personality lines, and login prompts
• Clean shutdown logic with sound effects

Tech Stack

Python, Tkinter, Pillow, pygame, playsound, requests, BeautifulSoup4, lxml,
TextBlob, OpenAI SDK, keyboard, colorama, winshell / pywin32, tkVideoPlayer

---

Project Structure

├─ src/ # source code │ ├─ JJ.py # entry point │ ├─ jazmin_application.py │ ├─ jazmin_userinterface.py │ ├─ jazmin_buttons.py │ ├─ jazmin_optimizer.py │ └─ jazmin_shortcut.py ├─ assets/ # runtime assets (images/sounds/fonts) ├─ docs/ │ └─ screenshots/ # images used in this README ├─ .env.example # placeholder env vars (no real keys) ├─ LICENSE (MIT) └─ README.md

Run Locally (optional)

The packaged installer is on the website. If you’d like to run from source:

# 1) Create & activate a venv (Windows)
python -m venv .venv
.venv\Scripts\activate

# 2) Install deps
pip install -r requirements.txt

# 3) Configure environment variables
copy .env.example .env   # then add your own keys

# 4) Launch
python src/JJ.py