This service prepares all public documentation for use with the large languge model. This process is called chunking and embedding. In this model it is assumes that the indee of the (regular) search engines passes each document to the chunker.