Question

A legal-tech team is building a RAG application to answer questions about lengthy commercial contracts. The documents contain deeply nested clauses, cross-references between sections, and formal defined terms that only make sense within their surrounding context. Early testing shows the assistant frequently returns partial answers that omit conditions stated in adjacent clauses. Which chunking strategy should the engineer adopt to best preserve semantic coherence for retrieval?

Accepted Answer

Use structure-aware (section/clause-based) chunking with overlap between adjacent chunks so related clauses and defined terms stay together

Answer

Split documents into fixed 256-token chunks with no overlap to maximize the number of retrievable units in the index

Answer

Store each entire contract as a single chunk so the full document is always retrieved as one unit

Answer

Split on individual sentences to create the smallest possible chunks for highly granular retrieval