🧠 Jarvis – Local Voice-Controlled AI Assistant

Jarvis is a voice-activated, conversational AI assistant powered by a local LLM (Qwen via Ollama). It listens for a wake word, processes spoken commands using a local language model with LangChain, and responds out loud via TTS. It supports tool-calling for dynamic functions like checking the current time.

🚀 Features

• 🗣 Voice-activated with wake word "Jarvis"
• 🧠 Local language model (Qwen 3 via Ollama)
• 🔧 Tool-calling with LangChain
• 🔊 Text-to-speech responses via pyttsx3
• 🌍 Example tool: Get the current time in a given city
• 🔐 Optional support for OpenAI API integration

▶️ How It Works (main.py)

1. Startup & local LLM Setup

   • Initializes a local Ollama model (qwen3:1.7b) via ChatOllama
   • Registers tools (get_time) using LangChain

2. Wake Word Listening

   • Listens via microphone (e.g., device_index=0)
   • If it hears the word "Jarvis", it enters "conversation mode"

3. Voice Command Handling

   • Records the user’s spoken command
   • Passes the command to the LLM, which may invoke tools
   • Responds using pyttsx3 text-to-speech (with optional custom voice)

4. Timeout

   • If the user is inactive for more than 30 seconds in conversation mode, it resets to wait for the wake word again.

🤖 How To Start Jarvis

1. Install Dependencies
   Make sure you have installed all required dependencies listed in requirements.txt:

   pip install -r requirements.txt
   

2. Set Up the Local Model
   Ensure you have the qwen3:1.7b model available in Ollama.

3. Run Jarvis
   Start the assistant by running:

   python main.py