hybrid-search

What problem does it solve?

Hybrid search solves the challenge of delivering high-quality results by merging semantic vector similarity with exact keyword matching.

Core Features & Use Cases

• Hybrid retrieval: merges vector similarity and BM25-like keyword matching using Reciprocal Rank Fusion for a unified ranking.
• Configurable weighting: adjustable alpha to balance semantic and lexical signals according to query type.
• Use Case: Build a search interface that understands intent and surfaces exact terms like product codes or model numbers.

Quick Start

Install llmemory and dependencies, then initialize a memory instance and execute a hybrid search with a sample query to observe blended results.