Summary on Knowledge Sources in Magnet AI
Knowledge sources are crucial for Retrieval-Augmented Generation (RAG) because they provide the foundational information that RAG uses to answer users' questions. Before setting up any RAG tool, you need to define the Knowledge Sources it will be working with.
To create a Knowledge Source, you will first need to define the source of content and then embed your content into the vector store by Syncing, which is done with just a button click. The supported content type is text (e.g. web pages, PDFs) or videos accompanied with a text transcript.