OpenSearch Project

@opensearch.org

OpenSearch is the trusted open source platform for AI-powered search, observability, and analytics with built-in security, high performance and a flexible architecture for modern applications. Learn more at https://opensearch.org/

Agent Health enables developers to combine deterministic checks with non-deterministic evaluations, using LLM-based judges to validate changed behaviors. Join Dotan Horovits and Megha Goyal as they explore how OpenSearch helps teams move beyond traditional observability: https://bit.ly/3Tv7aAI

OpenSearch 3.8 is here with up to 4x faster vector ingestion. Beyond faster vector workloads, the latest release expands MCP support across all agent types, improves radial search throughput by up to 2x, adds lower-latency ML streaming with gRPC, and more. Blog here: https://bit.ly/4fPSwf3

Bild

The Agent Skills Hackathon is live — and there's still time to jump in. OpenSearch Ambassador Neel Shah breaks down the details, why it's worth your weekend, and how to get started before submissions close. $5K in prizes! Submit your agent skills by August 17th! https://bit.ly/443K3iM

Agentic Relevance Tuning, presented at OpenSearchCon Europe 2026, brings LLM-powered agents into the search engineering workflow. ART helps detect relevance issues, generate tuning hypotheses, validate improvements, and automate the path toward better search results. Read: https://bit.ly/4bn4T0E

Bild

The OpenSearch Certified Practitioner (OSCP) certification is coming soon. Designed for beginners, OSCP will help practitioners validate their foundational OpenSearch knowledge and showcase their skills across the OpenSearch ecosystem. Be among the first to earn OSCP → https://bit.ly/45r7H9y

Bild

With agents capable of pushing 100K queries per min, search and observability need to scale it. EE Times features insights from Bianca Lewis on OpenSearch’s role in AI infrastructure, alongside Linux Foundation General Manager Arpit Joshipura on the evolution of the AI stack: https://bit.ly/3RW3PKE

Bild

When running gen AI and vector search in production, keeping ML endpoints off the public internet is essential for security. Nathalie Jonathan walks through how to configure OpenSearch's ML Commons plugin to connect to private endpoints within your internal network: https://bit.ly/3TJdyV4

Bild

OpenSearchCon NA 2026 has announced its full schedule. Join developers, architects, and enterprise leaders September 22-24 in San Jose for 60+ keynotes, lightning talks, breakout sessions, and hands-on workshops. Dive into building a complete OpenSearch observability stack: https://bit.ly/4wZlZdj

Bild

Karsten Schnitter from member company SAP explores how the OpenSearch KEDA scaler, contributed by SAP and available in KEDA 2.20+, enables teams to scale based on query results from logs, traces, and observability data without building a separate metrics pipeline. Read on: https://bit.ly/4vHr2Ow

Bild