DocGPT is designed to simplify information extraction and Q&A from documents by providing a seamless conversational interface. Users can upload documents in PDF, Word, or PowerPoint formats, which are then processed using text parsers. The content is chunked and embedded with OpenAI's embedding models, stored in a vector database like FAISS or Pinecone. When a user submits a query, DocGPT retrieves the most relevant text chunks via similarity search and leverages ChatGPT to generate accurate, context-aware responses. It features interactive chat, document summarization, customizable prompts for domain-specific needs, and is built on Python with a Streamlit UI for easy deployment and extensibility.