Elastic at AWS re:Invent: Concluding a year of partnership in agentic AI innovation

The integration of Elastic’s capabilities, including vector databases and context engineering, with AWS services helps customers build intelligent, scalable, and secure applications faster and with greater flexibility. Our ongoing collaboration has resulted in another year of notable innovation with AWS. This blog highlights our continued collaboration with AWS throughout 2025 to help you capitalize on the power of AI.

The Elasticsearch MCP Server connects your data to AI agents via the Model Context Protocol (MCP), enabling natural language interaction with your indices. It exposes Elasticsearch indices as secure, searchable resources, allowing large language models (LLMs) to perform powerful retrieval and analysis.

We also released a new integration for Amazon Bedrock AgentCore that brings observability for AI agents and applications directly into the Elasticsearch Platform. By ingesting logs and metrics, it helps you see performance, control costs, understand agent behavior, and keep workloads running reliably. This provides an end-to-end, unified view for deployments using Amazon Bedrock AgentCore and for monitoring LLM interactions across an organization’s agents.

Additionally, we created a guide about how to create an agent with Elastic Agent Builder and then explore how to use the agent via the A2A protocol orchestrated with the Strands Agents SDK.

The evolution of AI and its landscape has progressed at a breakneck pace. Our customers employing Elastic's solutions on AWS are benefiting from the scalability and agility of the cloud while maintaining the highest levels of speed and innovation with agentic AI. This progress was recognized by AWS through an advanced specialization in agentic AI.

As part of the AWS AI Competency, the Agentic AI Specialization represents the highest level of validation for partners working with autonomous AI systems. Partners undergo rigorous technical assessment and must demonstrate successful customer implementations that align with AWS’s commitment to responsible AI development. We were delighted to be among the first partners to receive the AWS Agentic AI Specialization.

We were thankful to be recognized as AWS partner of the year finalists in the following categories:

• Global - GenAI Infrastructure and Data

• EMEA - Technology

Elastic received a specific mention by Matt Yanchyshyn, vice president of AWS Marketplace & Partner Services, during his keynote speech for being at the leading edge of AWS partners providing agentic AI solutions to customers.

We were also highlighted by David Brown, vice president of AWS Compute and Machine Learning (ML) Services, during his keynote speech as one of the initial AWS partners for Graviton.

One of the main themes at AWS re:Invent was agentic AI. Elastic held a session to cover how developers can use Elasticsearch, Amazon Bedrock AgentCore, and Tavily to retrieve external content and accelerate the creation of AI agents. These agents provide more relevant results by leveraging context engineering capabilities native to open source Elasticsearch, such as semantic search, hybrid search, and ES|QL. 

We had a trifecta of speakers from Elastic (Uday Theepireddy, senior principal solution architect), AWS (Srinivas Pendyala, senior cloud solutions architect), and Tavily, an Elastic customer (Rotem Weiss, Founder and CEO).

These experts reviewed the underlying vector database capabilities that power these tools, recommended best practices for agent prompt design, evaluated agent performance, and created an intelligent query layer exposed through MCP. Check out the recording: Breakout Session: Context engineering and building better agents.

Logs carry the richest context for troubleshooting, yet brittle pipelines, fragile parsers, and cost-driven sampling keep teams stuck in reactive grep mode. In his session, David Hope, director of product marketing, introduced Elastic's AI-powered Streams, an agentic system that automatically understands and enriches any log format (structured or unstructured); simplifies ingestion, parsing, storage, and analysis; and surfaces significant events and anomalies out of the box. 

The result is that logs become your first and smartest investigative signal, accelerating root-cause analysis and delivering actionable insights in production. Watch the recording: Breakout Session: Live logs & prosper: Use AI to make logs a primary observability signal (AIM346).

Finally, Steve Kearns, general manager of search at Elastic, was joined by Ayan Ray, senior architect of generative AI at AWS, as part of AWS AI LIVE!. They spoke about building AI agents using context engineering, the importance of relevance and scalability in vector search, and Elastic Agent Builder.

This discussion mirrored that of Mike Nichols, general manager of security at Elastic, who joined AWS Security LIVE! in July at the AWS Partner Summit in NYC to provide Elastic’s perspective on challenges facing the security industry.

We invited several of our partners to present at the Elastic booth in the Expo hall. In addition to Accenture’s overview of the Data Readiness Engine, Gigamon presented GenAI insights with ML/AIOps. NVIDIA demonstrated GPU-accelerated vector search in Elasticsearch using NVIDIA cuVS, achieving up to 12x faster vector indexing. At the Confluent booth, we highlighted Elastic retrieval combined with Confluent streaming for real-time scale.

In case you missed them, we amplified announcements and encouraged event attendees to visit the Elastic booth via a series of posts on LinkedIn. We even had Elky join us!

• Elastic, Accenture, and AWS - Data Readiness Engine for GenAI

• Show floor buzz

• Elastic’s AWS AI Specialization in Agentic AI announcement and booth boost

• Integration of NVIDIA cuVS directly into Elasticsearch

• Elastic highlighted during AWS Marketplace Partner Advantage session

• AWS re:Invent recap

The release and timing of any features or functionality described in this post remain at Elastic's sole discretion. Any features or functionality not currently available may not be delivered on time or at all.

In this blog post, we may have used or referred to third party generative AI tools, which are owned and operated by their respective owners. Elastic does not have any control over the third party tools and we have no responsibility or liability for their content, operation or use, nor for any loss or damage that may arise from your use of such tools. Please exercise caution when using AI tools with personal, sensitive or confidential information. Any data you submit may be used for AI training or other purposes. There is no guarantee that information you provide will be kept secure or confidential. You should familiarize yourself with the privacy practices and terms of use of any generative AI tools prior to use. 

Elastic, Elasticsearch, and associated marks are trademarks, logos, or registered trademarks of Elasticsearch B.V. in the United States and other countries. All other company and product names are trademarks, logos, or registered trademarks of their respective owners.