OpenSearchCon North America 2026 brings together the teams building AI-ready search, observability, and security.
A lot has changed since the OpenSearch Project debuted in 2021.
AI agents are now querying data, calling tools, and making decisions inside production systems. That raises the bar for the infrastructure underneath them: search that can reason, not just match keywords, observability that can trace what an agent did and why, and security that can catch a threat before it becomes an incident. OpenSearch has grown into the platform many teams turn to for exactly that work, backed by a global community of more than 10,000 strong, 19 member organizations including AWS, IBM, SAP, and Uber, and more than 2 billion cumulative downloads.
But even more impressive than the size of the OpenSearch community is the range of problems it’s now solving. Enterprise search remains foundational to the project, but organizations increasingly rely on OpenSearch for observability, analytics, and vector search too.
This evolution will be on full display at OpenSearchCon North America 2026, in San Jose, Calif., from September 22 to 24. The conference marks five years of OpenSearch, and we see the milestone as an opportunity to rethink how the event can help our growing community.
The current moment is a turning point for OpenSearch, not a period of incremental change. What started as a search and analytics engine has scaled into the backbone for AI, observability, and security at some of the world’s largest organizations, and this year’s OpenSearchCon reflects that shift toward production-grade, enterprise-scale deployments.
The Critical Role of Search in AI Infrastructure
OpenSearch remains a foundational component of many organizations’ traditional enterprise search solutions, powering everything from in-house search tools for employees to retailers’ customer-facing websites. But increasingly, OpenSearch is playing an important role in AI solutions, as well. According to a new Linux Foundation Research report, which will be unveiled in full at the conference, 77% of organizations already consider OpenSearch a core or supporting component of their AI infrastructure.
OpenSearchCon 2026 will provide attendees with a wide range of real-world implementation strategies for generative and agentic AI workflows.
Another session, a panel called “Search, Agents, and Beyond: Customers Rewriting the Playbook on AI with OpenSearch,” features leaders from AWS, DTEX, and Airbnb, who will share how they’ve built agentic AI workflows where agents query, reason, and act on data in real time. This is an opportunity for attendees to go beyond the hype surrounding AI and learn what really works (and what doesn’t) in the real world.
At billion-document scale, the same principles show up in production. In “Hybrid Dense-Sparse Search at Billion Scale: Tuning HNSW + BM25 for Real-Time E-commerce Relevance,” Yogesh Sardana of Airtel will walk through how the telecom and retail giant tunes hybrid search to hit real-time relevance targets at massive scale.
These sessions and others illustrate how enterprise search is becoming more and more important as AI becomes more central to how organizations operate.
Even before generative AI tools became mainstream several years ago, teams struggled to organize and help users quickly access the enormous amounts of information they store. Generative and agentic AI have made that challenge even more consequential. And the organizations that overcome it will give themselves a significant edge in the marketplace.
AI Agents Raise the Stakes for Observability
Although OpenSearch began as a search and analytics engine, IT leaders soon realized its core capabilities, ingestion, indexing, and analysis, could be harnessed for organizational observability. Today, teams are expanding this role to correlate logs, metrics, and behaviors specifically linked to AI agents.
Observability has always been vital for understanding complex IT environments, but the rise of agentic AI makes it more critical than ever. Many organizations are only beginning to experiment with AI agents, and leaders are still discovering how these agents behave, the actions they can take, and the risks of unauthorized data access. Maintaining real-time monitoring is now essential for protecting sensitive information and staying in control.
It’s no longer enough for teams to know how long a service took to respond or which error it returned. They now need to understand what an AI agent did, which tools it called, and how requests moved between agents and services.
A hands-on workshop, titled “Pipe Dreams to Pipelines: Observing and Evaluating AI Agents and Services with OTel + OpenSearch,” will guide attendees through building a full OpenSearch + OpenTelemetry stack, including metrics, traces, and logs, all in a single session. The session, co-led by Anirudha Jadhav, senior engineering manager for OpenSearch at Amazon Web Services, and Wolfgang Theilmann, development manager at SAP, will cover how IT professionals can construct an agent’s reasoning path, correlate it against service metrics, and catch regressions in correctness, tool selection, cost, and latency.
That same discipline is already running in production. In “Monitoring OpenSearch Clusters at Scale: A Production-Grade, Open-Source Approach,” Sokratis Papadopoulos of CERN will show how his team monitors OpenSearch clusters supporting one of the world’s largest scientific research organizations.
As AI agents take on more prominent roles within organizations, observability practices must expand beyond infrastructure health. OpenSearchCon will give practitioners a chance to explore how OpenSearch and OpenTelemetry can provide the visibility they need as their organizations move agentic AI from experimentation into production.
Powering Active Threat Detection
For many organizations, OpenSearch is already sitting on top of large volumes of operational and security data. But many are currently missing out on opportunities to do more with this information than retaining it for observability dashboards and after-the-fact investigations. Security teams can also use OpenSearch’s analytics capabilities to identify anomalies and suspicious patterns in real time, laying a foundation for active threat detection. This shift is especially important as threats become more difficult to detect with static rules alone. “Dead Logs Don’t Lie: Turning Passive OpenSearch Log Storage into an Active Threat Intelligence” will show IT and security leaders how to transform “vanilla” log ingestion setups into an active threat intelligence engine, in the span of just 20 minutes. Akash Chadhokar, a senior software engineer at Tech Mahindra, will guide attendees through steps to activate OpenSearch’s anomaly detection logs to catch credential-stuffing attacks, cluster novel signature errors that precede system compromise, and wire everything into a real-time notification channel, all without an external SIEM, new infrastructure, or additional budget.
Some sessions will address security alongside observability and AI. For example, the “Search, Agents, and Beyond” panel mentioned above will explore how teams are using OpenSearch for intelligent log analysis where AI surfaces anomalies, as well as for security analytics in cases where threat detection moves from static rules to AI-powered correlation across logs, traces, and security events.
What to Expect in San Jose
This year’s keynote lineup spans the industry, with speakers from AWS, Apple, CERN, Intel, IBM, Aiven, Portal26, Uber, and SAP. Plus, the conference features session speakers from engineering teams at Airbnb, Airtel, Autodesk, Booking.com, Groupon, LinkedIn, NetApp Instaclustr, Uber, Walmart, and more.
Attendees will experience:
- Real-world implementation strategies for vector search, agentic AI workflows, and observability pipelines
- A look at how OpenSearch handles high-performance search and analytics at scale
- Sessions on search relevance tuning, plugin development, AI-powered observability, and vector database optimization
- Direct connections with the search engineers, architects, and open-source contributors building the next generation of AI-ready search systems
- Hands-on workshops where participants can build solutions alongside the people architecting solutions in the field
Five years ago, the question was whether OpenSearch could become a stable, production-ready project with a thriving community behind it. That question has been answered with a resounding “yes.” The questions facing the community today are bigger: how search, observability, and security keep up as AI agents take on more responsibility inside production systems.
The OpenSearch Project turns five this year, but OpenSearchCon 2026 isn’t a celebration of the past. Instead, it’s an opportunity for practitioners to come together and prepare for what comes next.
Learn more and register today at OpenSearchCon 2026. We look forward to seeing you there!