Blog
Building the future of OpenSearch together
As announced today, the OpenSearch Project is now part of the newly formed OpenSearch Software Foundation, a community-driven initiative under the Linux Foundation. This marks a major milestone in the history of the OpenSearch Project, and we couldn't be more excited to share our thoughts on what this means to...
Reducing hybrid query latency in OpenSearch 3.1 with efficient score collection
Hybrid queries combine the precision of traditional lexical search with the semantic power of vector search. In OpenSearch 3.1, we delivered significant latency reductions for hybrid queries by redesigning how...
Neural sparse models are now available in Hugging Face Sentence Transformers
OpenSearch’s neural sparse search transforms text into sparse token-weight pairs using transformer models, combining semantic search capabilities with efficient inverted indexing. This approach delivers high retrieval accuracy with low latency...
Enhanced log analysis with OpenSearch PPL: Introducing lookup, join, and subsearch
OpenSearch 3.0 introduced powerful new capabilities to the Piped Processing Language (PPL) with the addition of the lookup, join, and subsearch commands. These commands help you enrich, correlate, and filter...
Optimizing inference processors for cost efficiency and performance
Inference processors, such as text_embedding, text_image_embedding, and sparse_encoding, enable the generation of vector embeddings during document ingestion or updates. Today, these processors invoke model inference every time a document is...
Do more with less: Save up to 3x on storage with derived vector source
If you’re working with modern applications, from semantic search to recommendation systems, you’re likely implementing vector search. While you might focus on the accuracy and speed of vector similarity searches,...
Finding a replacement for JSM in OpenSearch 3.0
OpenSearch 3.0.0 introduced many innovative features that provide significant advancements in performance, data management, vector database functionality, and more. In the release announcement, we shared that OpenSearch has replaced the...
Intelligent troubleshooting using OpenSearch 3.0's plan-execute-reflect agent
OpenSearch 3.0 introduces the plan–execute–reflect agent—a powerful new capability that breaks down complex problems, selects and executes tools autonomously, and adapts through reflection. In this post, we’ll show you how...