ask-multiple-pdfs

Why we included this project

This is a small Streamlit app that shows the whole retrieval pipeline in one place. You load several PDFs, the text is split into chunks and embedded, and questions get answered by pulling the most relevant passages and handing them to a language model. Because everything lives in a single Python file, it is easy to follow how loading, chunking, embedding, and similarity search fit together, which makes it a good teaching example for anyone new to retrieval-augmented question answering. It also works as a real starting point for a document Q&A prototype, and the code supports both OpenAI and Hugging Face embeddings, so you can swap providers without much rewriting. If you want to understand the mechanics of asking questions of your own documents before building something more production-grade, this is a useful place to start.