Intelligent Search
A hybrid search engine that combines semantic understanding with keyword matching. Find results by meaning when you cannot remember exact words, or by precise terms when you need exact matches. The best of both approaches in a single search bar.
How it works
Every search query is processed through two parallel paths: semantic search converts your query into a meaning vector and finds conceptually similar content, while keyword search matches exact terms and phrases. Results from both are merged and ranked by a fusion algorithm that weighs relevance signals from each approach. Filters for date, type, vault, and tags narrow results further.
Search results showing semantic and keyword matches ranked by relevance
Why it matters
Pure keyword search fails when you cannot remember the exact words. Pure semantic search can miss when you need a specific term. Hybrid search handles both gracefully -- search for 'how to handle auth' and find items about 'authentication,' 'authorization,' 'login flow,' and 'JWT tokens' alongside exact matches. You find what you need regardless of how you phrase the query.
Search bar with type-ahead suggestions from both semantic and keyword engines