The RAG Voice Chatbot is an interactive voice-enabled Retrieval-Augmented Generation (RAG) assistant built using:
- Streamlit for the web interface
- Groq Whisper for speech-to-text transcription
- OpenAI GPT-4o for contextual question answering
- Pinecone for semantic vector storage and retrieval
- Groq TTS for text-to-speech answers
With this app, users can ask questions by speaking, and the chatbot will understand, retrieve relevant info from PDFs, and respond with a natural-sounding voice.
✅ Voice input using streamlit-mic-recorder
✅ real time streaming
✅ PDF document ingestion and vector embedding
✅ Context retrieval via Pinecone semantic search
✅ Smart answers powered by GPT-4o
✅ Record functionality.
✅ Real-time speech responses (TTS)
✅ Persistent chat history with audio playback
✅ Modern Streamlit UI with themed styling
Or see the screenshot below 👇
git clone https://github.com/yourusername/rag-voice-chatbot.git cd rag-voice-chatbot