Voice RAG Assistant

This project combines voice interaction with Retrieval-Augmented Generation (RAG) to create an assistant that can answer questions from custom documents and knowledge sources. Users can ask questions through speech, the system retrieves the most relevant context from a vector database, and the assistant responds with a grounded answer through voice.

The project is designed for use cases where users need fast, conversational access to private or specialized information, such as internal documentation, reports, support material, research notes, or business knowledge bases. The system can be expanded with better retrieval pipelines, source citations, multi-document support, and production-ready voice interfaces.