https://github.com/sameerqureshii/jarvis-ai-assistant

A voice controlled AI assistant built with Python that combines speech recognition, text-to-speech, and Google's Gemini AI API to create a personalized virtual assistant experience.
https://github.com/sameerqureshii/jarvis-ai-assistant

Last synced: 5 months ago
JSON representation

A voice controlled AI assistant built with Python that combines speech recognition, text-to-speech, and Google's Gemini AI API to create a personalized virtual assistant experience.

• Host: GitHub
• URL: https://github.com/sameerqureshii/jarvis-ai-assistant
• Owner: sameerqureshii
• License: mit
• Created: 2025-08-12T14:55:33.000Z (11 months ago)
• Default Branch: main
• Last Pushed: 2025-08-12T15:54:46.000Z (11 months ago)
• Last Synced: 2025-08-12T17:18:35.066Z (11 months ago)
• Size: 15.6 KB
• Stars: 1
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# Jarvis AI Assistant 

A voice controlled AI assistant built with Python that combines speech recognition, text-to-speech, and Google's Gemini AI to create a personalized virtual assistant experience.

# ✨ Features:

🎤 `Voice Recognition`: Wake word detection with "Jarvis"


🗣️ `Text-to-Speech`: Natural voice responses using Google TTS


🧠 `AI Integration`: Powered by Google Gemini for intelligent responses


🌐 `Web Navigation`: Quick access to popular websites


🎵 `Music Playback`: Voice controlled music from your library


📰 `News Updates`: Fetch latest headlines from News API


⚡ `Fast Commands`: Optimized hardcoded commands for common tasks


# 🛠️ Installation:

**Prerequisites:**

  1. Python 3.7 or higher


  2. Requires a microphone for voice input


  3. API keys must be valid for features to work


  4. Requires an internet connection for gTTS, NewsAPI, and Gemini AI


# Required Dependencies:

pip install `speechrecognition`


pip install `pyttsx3`


pip install `requests`


pip install `google-generativeai`


pip install `gtts`


pip install `pygame`


pip install `pyaudio`


# API Keys Setup:

  • News API: Get your free API key from NewsAPI.


  • Google Gemini API: Get your API key from Google AI Studio.

# 📁 Project Structure:

jarvis-ai-assistant/


├── main.py           # Main assistant code


├── musicLibrary.py   # Predefined song names & YouTube links


└── README.md         # Project documentation

# ⚙️ Configuration:

Before running the assistant, update the API keys in `main.py`:

  • NEWS_API_KEY = "Your-API"


  • GEMINI_API_KEY = "Your-API"

# 🚀 Usage:

**Run the assistant:**

python `main.py`

  1. Wait for "Initializing Jarvis..." message

  2. Say "Jarvis" to wake the assistant

  3. Wait for "Yes, Sameer" response

  4. Give your command within 6 seconds

# 📋 Function Documentation:

**Core Functions:**

`speak(text)`

**Purpose:** Converts text to speech using Google's TTS engine.

**How it works:**

  1. Creates a gTTS (Google Text-to-Speech) object

  2. Saves audio to temporary MP3 file

  3. Uses pygame to play the audio

  4. Automatically cleans up temporary files

  5. Falls back to error handling if TTS fails

**Parameters:**

  • `text` (string): Text to be spoken

`speak_old(text)`

**Purpose:** Alternative text-to-speech using offline pyttsx3 engine.

**How it works:**

  • Uses the local pyttsx3 engine for offline TTS


  • More reliable but less natural sounding than Google TTS

**Parameters:**

  • `text` (string): Text to be spoken

`aiProcess(command)`

**Purpose:** Processes natural language commands using Google Gemini AI.

**How it works:**

  1. Sends user command to Gemini 1.5 Flash model

  2. Uses system instruction to maintain Jarvis personality

  3. Returns AI generated response text

  4. Includes error handling for API failures

**Parameters:**

  1. `command` (string): User's voice command

**Returns:** AI generated response text

`processCommand(c)`

**Purpose:** Main command processor that handles both hardcoded and AI commands.

# How it works:

**1. Website Navigation (Hardcoded for speed)**:

  • "open google" → Opens Google


  • "open chatgpt" → Opens ChatGPT


  • "open github" → Opens developer's GitHub profile


  • "open linkedin" → Opens developer's LinkedIn profile

**4. Music Playback:**

  • Command format: "play [song_name]"


  • Searches musicLibrary.py for song links


  • Opens YouTube links in default browser


  • Provides feedback if song not found

**3. News Updates:**

  • Triggers on "news" keyword


  • Fetches top US headlines from News API


  • Reads first 3 headlines aloud


  • Handles API errors gracefully

**5. AI Fallback:**

  • Any unrecognized command goes to Gemini AI


  • Provides intelligent responses for general queries


**Parameters:**

  • `c` (string): Voice command from user

# 🎵 Music Library:

The `musicLibrary.py` file contains a dictionary of song names mapped to YouTube URLs:

music = {


    "stealth": "YouTube URL",


    "march": "YouTube URL",


    "skyfall": "YouTube URL",


    "wolf": "YouTube URL"


}

**Usage:** Say "play [song_name]" where song_name matches a key in the dictionary.


**Adding Songs:** Add new entries to the music dictionary in the format `"song_name": "youtube_url"`

# Modifying Voice Settings:

Adjust recognition parameters in the main loop:

  • `timeout`: Maximum wait time for voice input


  • `phrase_time_limit`: Maximum length of spoken phrase


  • `duration`: Ambient noise adjustment time

# Changing AI Personality:

Modify the system instruction in the Gemini model configuration:

`system_instruction="Your custom personality instructions here"`

# 📄 License:

This project is open source and available under the MIT License.

# 🤝 Contributing:

Feel free to fork this project and submit pull requests for improvements!

# 👨‍💻 Developer:

**Sameer Ahmed Qureshi**

`Portfolio`: https://sameer-personall-portfolio.vercel.app 


`LinkedIn`: https://www.linkedin.com/in/sameer-ahmed-qureshi56



Built with using Python and Google AI