Vector search and embeddings
View All Tags
You search for something you know is there, and it does not come back, because the writer used different words. We ran that exact failure against nine embedding configurations from three vendors and published every number. Three of them ranked the wrong document first.
Lucenia 0.8.1 delivers major performance gains with faster search, improved vectors, and stronger retrieval foundations.
Read More
Lucenia 0.5.0 introduces secure ML training, BYOM support, native RAG processors, and 80% smaller vector storage.
Read More
Lucenia 0.4.0 introduces KNN vector search enhancements, Lucene 10.1 upgrade, and expanded geospatial CRS capabilities.
Read More
Examining the intersection of AI and vector search - cutting through the hype to explore practical applications.
Read More
Key takeaways from the 2024 GEOINT Symposium on hybrid cloud, storage costs, and vector data for AI.
Read More
Balancing vector storage efficiency with search quality through compression and dimensionality reduction.
Read More
Understanding the curse of dimensionality and how Lucenia uses dimensionality reduction to cut storage costs.
Read More