Splitting documents for retrieval: fixed-size versus structural boundaries, overlap, what metadata to attach, and how to handle tables, code and long documents.https://contextaugmented.com/https://contextaugmented.com/blogs/chunking-tables-code-and-transcripts/https://contextaugmented.com/blogs/chunking-tables-code-and-transcripts/Three document types that break general splitters, and the handling each needs: repeated headers, syntax-aware cuts, and topic-grouped speaker turns.Sun, 26 Jul 2026 00:00:00 GMTdocument-typesboundarieshttps://contextaugmented.com/blogs/metadata-is-half-the-chunk/https://contextaugmented.com/blogs/metadata-is-half-the-chunk/A chunk with no source, date or heading path can't be filtered or cited. What to attach, what to embed with the text, and what to keep out of the vector.Sat, 25 Jul 2026 00:00:00 GMTmetadatastrategyhttps://contextaugmented.com/blogs/overlap-and-what-it-actually-buys/https://contextaugmented.com/blogs/overlap-and-what-it-actually-buys/Overlap insures against cutting through an idea. It also inflates your index and fills top-k with near-copies. When it earns its cost, and when it's a patch.Fri, 24 Jul 2026 00:00:00 GMToverlapstrategyhttps://contextaugmented.com/blogs/split-on-structure-not-on-characters/https://contextaugmented.com/blogs/split-on-structure-not-on-characters/A character splitter cuts where the counter runs out. Documents already have boundaries — headings, clauses, functions. Use those; count only as a fallback.Thu, 23 Jul 2026 00:00:00 GMTboundariesstrategyhttps://contextaugmented.com/blogs/why-chunk-size-is-a-tradeoff/https://contextaugmented.com/blogs/why-chunk-size-is-a-tradeoff/Small chunks retrieve precisely and answer incompletely. Large chunks do the reverse. What each end actually breaks, and how to find your corpus's middle.Wed, 22 Jul 2026 00:00:00 GMTchunk-sizestrategy