This search updates itself

Build reactive full-text search with Convex, from defining a search index to subscribing from the client so results update automatically while you type and whenever matching data changes. In this video, I build a searchable chat app and cover: - Defining a `searchIndex` with a search field and exact filter fields - Using nested fields with dot syntax - Staging an index while Convex backfills a large existing table - Querying the index with `withSearchIndex` - Subscribing with `useQuery` for automatic reactive updates - How multiple terms, case sensitivity and exact filters behave - Keyword search vs semantic/vector search - Tantivy, relevance ranking and BM25 - The practical limits you should understand before shipping Try the demo: https://demo-convex-full-text-search.vercel.app/ Source code: https://github.com/mikecann/demo-convex-full-text-search Full-text search docs: https://docs.convex.dev/search/text-search Current limits and pricing: https://docs.convex.dev/production/state/limits#full-text-search 00:00 Reactive search in action 01:17 Define the search index 02:27 Query the index 02:52 Subscribe from the frontend 03:24 Test the finished chat app 04:44 Case sensitivity and exact filters 05:06 Keyword search vs semantic search 05:23 How full-text search works under the hood 05:50 Relevance and result ranking 06:17 Combining full-text and vector search 06:35 Limits you need to know 07:27 Wrap-up #Convex #FullTextSearch #WebDevelopment