A multi-provider RAG application with a Tkinter GUI. It builds a local vector knowledge base from PDFs using FAISS and sentence-transformers. The app queries this data and sends prompts to various LLM APIs like Google, OpenAI, and Anthropic. A build script bundles the app and embedding model into a single executable for distribution.