Gyani listens, understands Hindi & English, thinks with GPT, and talks back — all hands-free.
Back in 2021, ChatGPT didn't exist. Asking your computer a question and getting a spoken, intelligent answer back was not something most people had seen outside of Siri or Alexa — both of which required expensive ecosystems to build on.
Gyani was built as a solo project to explore what was possible with the OpenAI API (which was barely public at the time), Python speech recognition, and desktop automation — all stitched together into a Hindi-English voice assistant that could genuinely hold a conversation, read you the news, control your PC, and more.
This was before the AI wave. Before everyone had a ChatGPT wrapper. Just curiosity, Python, and a mic. 🎤
This is the original 2021 source code, preserved as-is for archival purposes.
A lot has changed in the Python/AI ecosystem since 2021. If you try to run this code today, several things will likely break:
| What's broken | Why | Fix |
|---|---|---|
openai API calls |
OpenAI released v1.0 SDK in 2023 — completely different syntax | Upgrade to openai>=1.0 and rewrite calls |
googletrans |
Unofficial API breaks frequently | Pin to googletrans==3.1.0a0 |
pyaudio |
Needs PortAudio C library — painful on modern Windows | Use conda install pyaudio |
BeautifulSoup scraping |
Google changes its HTML structure often — temperature scraping may fail | Use a proper weather API instead |
| API keys | All hardcoded keys have been removed and are expired anyway | Get fresh keys from each provider |
| Python 3.14 | Many packages don't have wheels for 3.14 yet | Use Python 3.11 |
TL;DR — This code was written in 2021 with library versions from that era. It is being actively updated and refactored. Treat this as a reference / work-in-progress, not a plug-and-play project.
| Feature | Description |
|---|---|
| 🎤 Voice Input | Speaks in Hindi or English — Gyani understands both |
| 🌐 Auto Translation | Hindi speech auto-translated to English via Google Translate |
| 🧠 GPT-Powered Brain | Conversations powered by OpenAI GPT-3.5-turbo |
| 🔊 Text-to-Speech | Responds out loud using pyttsx3 TTS engine |
| 📰 Live News | Fetches top headlines by category via NewsAPI |
| 🌡️ Weather/Temperature | Scrapes real-time temperature from Google |
| 😂 Jokes | Tells random programmer jokes |
| 🎵 Spotify Control | Opens and plays Spotify Daily Mix automatically |
| ⚡ Speed Test | Checks your internet download & upload speed |
| 🖼️ Image Generation | Generates AI images using DALL-E 2 |
| ⏰ Reminders | Sets desktop notification reminders |
| 🎂 Birthday Alerts | Reads from Excel and announces today's birthdays |
| 🖥️ System Control | Shutdown, sleep, hibernate, lock, recycle bin |
| 👏 Wake Word / Clap | Wake Gyani with a clap or by saying "wake up" |
| 💬 Chat Logs | Every conversation saved to chatlogs.txt |
GyaniAssistant/
│
├── brain.py # Core AI engine (GPT-3.5 integration)
├── Listen.py # Microphone input + Hindi→English translation
├── Speak.py # Text-to-speech output (pyttsx3)
├── wishing.py # Time-based greeting (Good Morning/Afternoon/Evening)
├── clap.py # Wake detection (clap or voice "wake up")
│
├── Tell_News.py # Live news via NewsAPI
├── Temperature.py # Real-time weather scraping
├── Joke.py # Random jokes via pyjokes
├── spotify.py # Spotify automation via pyautogui
├── speed_test.py # Internet speed test
├── image_generator.py # DALL-E 2 image generation
├── SetReaminder.py # Desktop reminder notifications
├── birthdates.py # Birthday detection from Excel
├── systemRelated_info.py # System control (shutdown, lock, etc.)
├── song_detection.py # Song title detection from lyrics
├── gyani.py # Google Maps integration
│
├── gui.py # GUI (customtkinter) — in progress
├── dates.xlsx # Birthday data file
├── chatlogs.txt # Conversation history
├── requirements.txt # Python dependencies
└── Gyani.ico # App icon
- Python 3.11 (strongly recommended — not 3.12+ or 3.14)
- Windows OS (some modules are Windows-only)
- A working microphone
1. Clone the repository
git clone https://github.com/abhigyan24-ops/GyaniAssistant.git
cd GyaniAssistant2. Install dependencies
pip install -r requirements.txt
⚠️ PyAudio on Windows requires PortAudio. Ifpip install pyaudiofails:conda install pyaudio
3. Set up your API keys
Create a .env file in the root folder:
OPENAI_API_KEY=your_openai_key_here
NEWS_API_KEY=your_newsapi_key_here
GENIUS_API_KEY=your_genius_key_here
GOOGLE_MAPS_API_KEY=your_google_maps_key_hereThen replace the YOUR_OPENAI_API_KEY placeholders in the .py files with os.getenv("OPENAI_API_KEY").
4. Run Gyani
python clap.py # Start with wake detection
# OR
python brain.py # Start conversation directly| Service | Used In | Get Key |
|---|---|---|
| OpenAI | brain.py, image_generator.py |
platform.openai.com |
| NewsAPI | Tell_News.py |
newsapi.org |
| Genius | song_detection.py |
genius.com/api-clients |
| Google Maps | gyani.py |
console.cloud.google.com |
👏 Clap / Say "wake up"
↓
wishing.py → Good Morning / Afternoon / Evening
↓
Listen.py → Mic input → Hindi/English → Translated text
↓
brain.py → GPT-3.5 generates response
↓
Speak.py → Speaks the response aloud
↓
[Feature modules triggered by voice command]
News / Weather / Joke / Spotify / Reminder / etc.
This project is being modernised. Here's what's planned:
- Upgrade OpenAI SDK from v0 to v1.x
- Move all API keys to
.envwith python-dotenv - Fix
SetReminder.pyelif chain bug - Replace Google scraping in
Temperature.pywith a proper weather API - Build full GUI with customtkinter (
gui.py) - Add voice command intent detection to route to feature modules
- Cross-platform support (Linux / macOS)
Pull requests are welcome! Especially help with the refactor roadmap above.
- Fork the repo
- Create your branch (
git checkout -b feature/YourFeature) - Commit your changes (
git commit -m 'Add YourFeature') - Push to the branch (
git push origin feature/YourFeature) - Open a Pull Request
Abhigyan — @abhigyan24-ops
Built in 2021 as a solo project — when building something like this actually took some digging.
Made with ❤️ and a lot of speak("debugging...") calls
⭐ Star this repo if you found it useful!