🗣️ AI Voice Assistant with LangChain + Ollama

A simple voice assistant that listens to your speech, converts it to text, queries a local LLM using LangChain, and speaks the response back using TTS.

---

🎯 Features

• 🎤 Speech-to-text using SpeechRecognition + microphone
• 🤖 Local LLM querying via langchain-ollama (Mistral model)
• 🧠 Maintains chat history using LangChain memory
• 🔊 Text-to-speech using pyttsx3
• 🖥️ UI powered by Streamlit

---

📦 Installation

1. Clone the repo

git clone https://github.com/AzkaSahar/AI-Voice-Assistant.git
cd AI-Voice-Assistant

2. Install dependencies

pip install -r requirements.txt

⚠️ You may need to install pyaudio manually:

pip install pipwin
pipwin install pyaudio

---

🚀 Usage

1. Start Ollama (if not already running)

Make sure Ollama is installed and the mistral model is available:

ollama run mistral

2. Run the Streamlit app

streamlit run ai_voice_Assistant.py

---

🧠 How It Works

• Listens to user voice input
• Transcribes using SpeechRecognition
• Passes query and history to LangChain-powered Mistral model via langchain-ollama
• Speaks response using pyttsx3
• Maintains a visible chat history

---

📝 License

MIT – Free to use, modify, and share.