Qdrant, a leading high-performance open-source vector database, has launched BM42, a pure vector-based hybrid search approach for modern retrieval-augmented generation (RAG) applications. This new search algorithm combines traditional text search and vector search to provide more accurate and efficient retrieval, catering to the specific demands of RAG scenarios. By shifting from keyword-based search to a fully vector-based approach, Qdrant aims to set a new industry standard, offering a more flexible, precise, and efficient solution for short text segments.
BM42 introduces a new way of classifying search results, integrating sparse and dense vectors to accurately pinpoint relevant information within a document. This hybrid search system boosts accuracy, efficiency, and scalability, providing a cost-effective solution for developers. Qdrant’s BM42 is designed to enable users to quickly transition from prototype to production and scale the solution globally.
Qdrant is a high-performance, scalable, open-source vector database and search engine that supports billions of vectors and complex object matching. The company was recently recognized among the top 10 startups on Sifted’s 2024 B2B SaaS Rising 100 list. For more information about Qdrant and BM42, visit their website or contact press@qdrant.com.
Source link
Source link: https://www.01net.it/qdrant-launches-groundbreaking-pure-vector-based-hybrid-search-setting-higher-standards-for-rag-and-ai-applications/