openai-doc-reader is a RAG tool based on the OpenAI Assistant File Search & Vector Store API (V2) designed to read, analyze, and interact with documents. This project allows users to upload files (such as PDFs), transform them into embedding vectors, and then interact with these documents through an AI-powered chat interface.
- Document Upload and Processing: Allows users to upload files, which are then processed to extract embeddings.
- Advanced Text Analysis: Utilizes the OpenAI API to analyze the text of documents and provide contextual responses to questions.
- Interactive Chat Interface: Provides an intuitive user interface for asking questions about the document content and receiving real-time answers.
- Citations and References: Enriches responses with citations and references for each answer, making it easy to locate the source information.
-
Clone this repository :
git clone https://github.com/yantheo/openai-doc-reader.git
-
Navigate to the directory : cd openai-doc-reader
-
Install the dependencies : pip install -r requirements.txt
-
Configure the necessary environment variables by creating a .env file: : OPENAI_API_KEY=your_openai_api_key
-
You need also to create your own thread id and assistant before running it
-
Run the application : streamlit run app.py