Chunking
Splitting a long document or transcript into smaller pieces ("chunks") so they can be embedded and retrieved individually by a search or RAG system.
Chunking is the step that turns a long transcript into retrieval-sized pieces. Each chunk is embedded and stored separately so a query can pull back just the relevant passage rather than an entire hour-long recording.
Good chunking for audio respects natural boundaries — speaker turns or topic shifts — instead of cutting at a fixed character count, and carries metadata (speaker, timestamp, source) on every chunk so retrieved results stay citable.