🧠🎙️ Offline Voice Assistant (CSV + RAG + Qwen + Whisper Compatible)

This is a fully offline-capable voice assistant system that:

• 🎤 Accepts voice input using either:
  • Google Speech Recognition (online)
  • OpenAI Whisper (offline)
• 🔍 Uses FAISS + CSV + RAG (Retrieval-Augmented Generation) for knowledge-based Q&A
• 🧠 Runs a local LLM (e.g., Qwen 0.5B or compatible) to generate natural responses
• 🔊 Replies using offline Text-to-Speech via pyttsx3

---

🚀 Features

• ✅ Works completely offline
• ✅ Uses your own CSV files as knowledge base
• ✅ Supports Whisper for offline speech-to-text
• ✅ Integrates Qwen LLM for smart local reasoning
• ✅ Fast search using FAISS vector database
• ✅ Voice responses using pyttsx3 TTS engine

---

📦 Dependencies

Install the following Python packages in a virtual environment:

pip install whisper
pip install faiss-cpu
pip install pyttsx3
pip install transformers
pip install sentence-transformers
pip install pandas
pip install torch
pip install sounddevice
pip install scipy

##Architecture Overview

graph TD; MIC[🎤 Microphone Input] --> STT[🗣️ Speech-to-Text (Whisper)] STT --> QUERY[❓ User Query] QUERY --> RAG[🔍 RAG + FAISS Search] RAG --> LLM[🧠 Local LLM (Qwen)] LLM --> TTS[🔊 Text-to-Speech (pyttsx3)] TTS --> AUDIO[📢 Speak Output]

Folder Structure

offline_voice_assistant/ │ ├── data/ │ └── knowledge.csv # Your knowledge base │ ├── models/ │ └── qwen_model/ # Local LLM (Qwen or compatible) │ ├── audio_input/ │ └── user_input.wav # Captured audio │ ├── assistant.py # Main script ├── rag_faiss.py # RAG + FAISS logic ├── speech_to_text.py # Whisper-based STT ├── text_to_speech.py # pyttsx3-based TTS ├── utils.py # Utility functions └── README.md

##Sample CSV Format

id,question,answer 1,What is AI?,Artificial Intelligence is the simulation of human intelligence in machines. 2,What is Python?,Python is a popular programming language used for AI and automation. ...

###How to Run Make sure your model (e.g., Qwen) is downloaded locally.

Place your CSV in the data/ folder.

Run the main assistant:

python assistant.py