How wine.co.za uses Manticore Search alongside MariaDB to improve search across its wine industry archive, provide better suggestions, and keep maintenance to a minimum.
An embedding model reads only the first few hundred tokens of a document and silently drops the rest. Manticore Search now splits long documents for you at INSERT time: add chunk_strategy to the vector column and pick one of five strategies. No ingest pipeline, no splitter library. On our own manual, recall@5 for deep content went from 55% to 83%.
How Bookbot uses Manticore Search to keep 1.29 million second-hand books searchable, match new arrivals against 3.5 million saved searches, and power a channel responsible for ~10% of revenue.
Manticore Search 29.9.0 adds chunked and multi-vector auto-embeddings, embedding input limits, UTF-8 identifiers, mmap columnar access by default, and fixes for hybrid search, KNN, bulk ingestion, and grouping.
Connect OpenSearch Dashboards to Manticore Search, explore data in Discover, and build visualizations and dashboards using a familiar interface.
A practical example of how to select several most recent items from each group in Manticore Search and combine them into a single string.