Semantic Understanding + Precise Matching
Vector search understands the meaning of questions, while keyword search precisely matches specific terms. Get the best of both worlds.
Excellent for Technical Terms
Product names, internal terminology, and technical jargon that vector search alone might miss are reliably captured.
Improved Response Accuracy
Filter out irrelevant documents and reference only the most pertinent information to generate accurate responses.
How It Works
1
Vector Search
Converts the question into a numerical vector and calculates semantic similarity with registered content.
2
Keyword Search
Performs full-text search using important keywords in the question to extract relevant documents.
3
Score Fusion
Combines results from both searches and selects the most relevant documents to feed to the AI.