<!channel> 🚀📣 We’re excited to announce availability of the latest version of Typesense

v0.24.0

, publicly today! 🎉 🎊 Here are some key highlights: 👉 Native support for indexing nested objects and arrays of objects 👉 Cross-field OR and complex filter expressions 👉 Vector search: support for both exact & HNSW-based nearest-neighbor vector searching. This lets you build features like similarity search, recommendations, visual search and semantic search using embeddings from your machine learning models, indexed in Typesense. 👉 Multi-lingual support: Chinese, Japanese, Korean, Cyrillic and Thai are now supported via the use of field-level locale flag in the collection schema. 👉 Optional filtering:

sort_by

clause can now accept a boolean expression whose result is used for sorting, e.g.

sort_by=_eval(brand:nike):desc,_text_match:desc

will boost all products that have brand as nike to the top. You can conditionally trigger this using the Curation/Overrides API as well. 👉 Preset search configurations: manage your search query parameters by storing them within Typesense as a “preset”, that you can refer to during query time. This helps you change these parameters on the fly from inside Typesense, without having to deploy code each time. 👉 You can now setup time windows when overrides should automatically be enabled and then disabled. 👉 You can now trigger an override when a particular filter expression is used, in addition to the search query used. 👉 We’ve also improved memory usage by ~15%-20%, compared to v0.23.1 when we benchmarked with some of our test datasets. Read the full changelog here and instructions on how to upgrade here. As always, a big thank you to everyone who contributed code, ideas, feature requests, bug reports and feedback! This is a crucial component of how we shape and build Typesense, so I’m really appreciative of your support. 🙏 🙏 🙏