THE TEXTS WERE BROKEN INTO FRAGMENTS, TRANSFORMED INTO MATHEMATICAL REPRESENTATIONS, AND STORED IN A VECTOR DATABASE. A RETRIEVAL-AUGMENTED GENERATION (RAG) SYSTEM SEARCHES THIS ARCHIVE WITH EVERY QUESTION, FINDING THE PASSAGES MOST RELEVANT TO WHAT IS BEING ASKED. THE PRESENCE SPEAKS FROM THESE FRAGMENTS, BUT IS NEVER LIMITED TO THEM.