To allow the model to retrieve information from documents with in an easier way, we’re going to split these documents into smaller chunks that we will then embed into a vector database. you’ll thus need a document loader, a text splitter and an embedding model.