Chunk

Retrieval does not work over whole documents; it works over pieces. Chunking splits a long document into smaller passages, each indexed separately, so a search can return just the relevant part instead of a whole file. A chunk is one such passage.

How you chunk matters. Pieces that are too big bury the answer in noise; too small and they lose the context that makes them make sense. Good chunking is quiet, unglamorous work that decides how well everything downstream retrieves.