OpenSearch Project Blog

As vector search workloads scale, retrieving vectors from search results becomes a bottleneck. Vector retrieval using doc values in OpenSearch 3.7 bypasses the expensive _source vector reconstruction path, delivering 5.5x faster end-to-end retrieval and 14.7x faster server-side performance.

Deploy the OpenSearch Observability Stack—a single OpenTelemetry-native platform unifying metrics, traces, and logs—and follow a step-by-step root cause analysis from error spike to failing log line, without switching tools.

Learn how vector search in OpenSearch eliminates zero-result pages in e-commerce by acting as a semantic fallback when lexical query relaxing falls short.

The OpenSearch agent server is a multi-agent orchestration platform that routes tasks to specialized AI agents using MCP tools and intelligent context-aware routing.

Asymmetric embedding models in OpenSearch 3.5 improve semantic search relevance by up to 125% over symmetric models. Learn how they work and how to implement them.

Across all SDLC phases, AI agents now generate code, root-cause analyses, remediation plans, performance fixes, and standard operating procedures faster than humans can meaningfully review them. The bottleneck is no longer generation; it's deciding whether to trust what has been generated. This post shares four agents the OpenSearch team built across the SDLC and the harness-based verification loops that make them reliable.

Contact center customers shouldn't have to repeat themselves on every call. This post shows how OpenSearch agentic memory provides your AI agent with persistent memory across sessions. In this implementation, semantic and hybrid search retrieve relevant information on demand, so your agent remembers what matters.

The OpenSearch Hadoop connector 2.0 adds Apache Spark 3.5 and 4 support, OpenSearch 3.x compatibility, Amazon OpenSearch Serverless support, and AWS SDK v2 migration.

Documents with tables, charts, and diagrams need more than text-based search. We benchmarked three multimodal document search approaches in OpenSearch on 1,000 report pages. ColPali late interaction reranking achieved the highest search quality (92% recall) but with higher latency, BDA modality-aware embedding offered a balanced middle ground, and text-only chunking was simplest to implement but scored lowest on visually rich content.

Learn how OpenSearch 3.5's new context management feature solves token limit issues for AI agents through intelligent summarization, sliding windows, and smart truncation. Build long-running agents that maintain conversation continuity without context overflow.

OpenSearch now offers native gRPC endpoints with Protocol Buffers, featuring an automated conversion pipeline from OpenAPI specifications. Benchmarks show up to 17% throughput improvement for vector search and 22% latency reduction for bulk ingestion compared to REST.

Learn how agentic search in OpenSearch performs across search relevance and execution accuracy benchmarks, achieving 82% query translation accuracy and up to 235% relevance improvements.

OpenSearch doc value skip indexes deliver up to 28x faster aggregations on uncorrelated filter and aggregation fields, overcoming key filter rewrite limitations.

Discover why AI agents fail in silence and how OpenSearch Agent Health solves it with open-source trace observability, automated benchmarking, and LLM judge evaluation.

Discover how OpenSearch FP16 vector search performance improved up to 310% throughput using bulk SIMD in OpenSearch 3.5, with progressive optimizations across 3.1, 3.4, and 3.5 reducing p99 latency to 91 ms.

Explore OpenSearch 3.4's agentic search UI in Dashboards. Configure AI agents, test natural language queries, and export configurations with hands-on examples.

Discover how OpenSearch's Anomaly Detection handles missing data with configurable imputation, turning data silence into critical alert signals. Learn how configurable imputation strategies and Klarna's real-world monitoring use case.

Discover how random rotation in OpenSearch 3.2 improves binary quantization accuracy by removing axis correlations without changing true distances in your data.

Discover OpenSearch's experimental Apache Arrow integration. Learn how Arrow Flight enables zero-copy data transfer between nodes, improving throughput for large-scale analytics and observability workloads.

OpenSearch 3.3 introduces resource sharing and access control, enabling teams to securely share ML models, anomaly detectors, and other plugin resources with fine-grained permissions.