Elastic unveils Agent Builder for custom AI agents with real data

Elastic has launched Agent Builder, a developer tool designed to help organisations create custom AI agents that work directly with their proprietary data.

Agent Builder is embedded with conversational features and integrates deeply with Elasticsearch, Elastic's search platform, enabling users to develop and deploy AI agents that leverage data sourced from documents, emails, business applications, and feedback channels.

Operational lifecycle

The purpose of Agent Builder is to streamline what is known as context engineering, the process of ensuring AI agents access the right organisational data at the right time. Many enterprises contend with scattered and unstructured data sets, which creates obstacles to delivering accurate and contextually relevant responses during automated business processes.

Elastic describes Agent Builder as a comprehensive set of features that extends the capabilities traditionally found in Elasticsearch, making it simpler to manage agent development, configuration, execution, customisation, and monitoring within a single platform.

"AI agents don't just need lots of data, they need the right data and tools, with relevance, guardrails, and observability built in," said Ken Exner, Chief Product Officer at Elastic.

"Developers already rely on Elasticsearch to find the right answer from their messy business data. Agent Builder goes further by making Elasticsearch one of the fastest platforms to build precise AI agents that use your data, where retrieval, governance, and orchestration all operate in one place, natively."

Conversational access to data

Agent Builder includes a native conversational agent. This means developers and business users can ask questions of company data using natural language and receive direct responses, rather than relying on structured database queries. The conversational approach extends to configuring searches, selecting relevant data indexes, and tailoring agent parameters to specific organisational needs.

Elastic notes that Agent Builder incorporates intelligent search capabilities that automatically determine which indexes to query, interpret the data's structure, and optimise the way questions are translated for the underlying language model, ensuring only the most relevant context is returned. This is designed to improve the reliability and usefulness of AI outputs while maintaining the security and compliance of enterprise data.

Customisation and integration

With Agent Builder, developers can define custom tools for their AI agents, making use of Elasticsearch's own query language (ES|QL) to specify what data should be leveraged for context generation. This feature provides granular control over what the AI agent can access, thereby enhancing both accuracy and data security.

Agent Builder also allows users to establish bespoke agents tailored from the ground up. Developers have the option to customise the system prompt, adjust tool access, and configure security profiles to align with company policies and specific use cases.

For organisations operating with multiple AI tools and standards, Agent Builder can connect with external agents and applications via the Model Context Protocol (MCP) and Agent-to-Agent (A2A) connections, all while retention of governance is enforced through the Elasticsearch execution layer.

Feature overview

Elastic has highlighted several core capabilities available in Agent Builder, including:

• Immediate chat functionality with enterprise data inside Elasticsearch, providing the ability to converse with company information in real-time.
• Built-in intelligent search tools designed to enhance relevance by choosing the correct indexes and types of queries.
• Custom tool development using ES|QL for precise data control.
• Bespoke AI agent creation, allowing for persona and access configuration as well as security profiling.
• Safe integration with external agents and applications via MCP and A2A standards, ensuring ongoing governance within Elasticsearch.

Availability

Agent Builder is currently offered in Technical Preview on Elastic Cloud for serverless deployments and is expected to be available in version 9.2 in the near future.