Couchbase launches AI Data Plane for enterprise agents

Couchbase has launched its AI Data Plane for enterprise AI agents, its first major technology release since returning to private ownership.

The offering is designed as a unified data infrastructure layer to support AI agents with persistent memory, context retrieval and data access across cloud, edge and self-managed environments. It combines functions often deployed as separate tools, including agent memory, an agent catalogue and a self-managed MCP server for Model Context Protocol integration.

The launch also includes Enterprise Analytics 2.2, which extends Couchbase's analytics tooling to work with Apache Iceberg-based lakehouse environments. A Trino adapter is also planned, allowing SQL access to Couchbase operational data from Trino-based platforms without first moving that data into separate analytical systems.

Couchbase is framing the release around a problem many companies face as they try to move AI systems from early testing into day-to-day operations. According to the company and outside analysts, the challenge is less about the underlying models and more about how data is stored, retrieved and shared across applications and environments.

Devin Pratt, Research Director, AI, Automation, Data & Analytics at IDC, linked that challenge to the need for real-time access to context.

"Most enterprises quickly discover that moving from chat-style pilots to production-grade agentic systems is really a data problem, not just a model problem," Pratt said. "IDC expects that 80% of agentic AI use cases will require real-time, contextual, and widely accessible data, so the architecture has to support that. Approaches that make agent memory and context retrieval first-class capabilities of the database itself, like Couchbase's AI Data Plane, address this directly. By unifying vectors, documents, cache, and operational data in a single distributed platform, from cloud to edge, Couchbase reduces the integration tax that has been slowing down real-world agent deployments and gives organizations a more governable, scalable foundation for the next wave of AI-powered applications."

Data layer

At the centre of the new product is what Couchbase calls Agent Memory, which stores conversational context and session state alongside operational data. The approach is intended to let developers avoid maintaining separate vector, document and caching systems as they build applications that rely on AI agents.

Couchbase says the memory layer is framework-agnostic and has been validated with LangGraph, CrewAI and LlamaIndex. That means engineering teams can switch between orchestration frameworks, or use more than one, without rebuilding the underlying data layer.

The company also argues that edge deployments are becoming more important as AI agents are used in field operations, mobile applications and disconnected environments. In response, it has extended the AI Data Plane to mobile and edge systems so agents can use replicated data and local vector search even when connectivity is limited.

Updates in this area include Couchbase Lite 4.1, Edge Server 1.1, React Native 1.1 and Sync Gateway 4.1. Changes include peer-to-peer synchronisation over Bluetooth with switching to Wi-Fi, expanded support for Windows and Arm systems, and updates aimed at managing permissions and software updates across distributed device fleets.

Analytics link

Enterprise Analytics 2.2 is aimed at companies that want to query operational data alongside data already held in lakehouse systems built on Apache Iceberg. The release supports federation rather than extraction, allowing users to query across systems without duplicating data through separate ETL pipelines.

Other additions include support for Google Cloud Storage, JWT authentication, Oracle and SQL Server change data capture, asynchronous long-running queries, an index advisor, index-only query plans and SQL++ UPDATE support. Corresponding SDK updates are available for Java, .NET, Python, JavaScript and Go.

Couchbase has also updated Capella iQ, its natural-language query assistant, to support model selection across AWS Bedrock and OpenAI under organisation-level policies. Administrators can set which models are available to particular teams to manage costs and data residency requirements.

One customer already using Couchbase in conversational AI described data retrieval as the central technical issue.

"What matters most for enterprise-grade conversational AI agents is that data retrieval is fast, consistent, and seamless. When you're running human-to-AI agent interactions, everything behind the scenes needs to be predictable and consistent to provide natural interaction," said Patrick Ferriter, SVP of Product at Agora. "That's what we're solving together with Couchbase, and it's why we chose them as a partner for the data layer for our conversational AI platform. Every one of our conversational AI use cases requires efficient data retrieval to feed the pipeline for AI agents, whether that's outbound sales, customer service, physical AI, or something entirely new. We've had a multi-year relationship with Couchbase, and as we've scaled into agentic workloads, this was a natural extension to our partnership."

Strategic signal

The release is also notable as an early product move under Couchbase's reshaped leadership team following the company's take-private deal. That gives the launch added significance as an indication of where management wants to focus the product in a market where database vendors are trying to tie core infrastructure more closely to AI application development.

Barry Morris, Chief Product and Strategy Officer at Couchbase, said customers had pushed the company to reduce the number of systems needed to support production AI agents.

"The database layer is where agentic AI either scales or stalls, and most of the industry is still treating agent memory as an afterthought," Morris said. "We built the AI Data Plane because our customers told us that stitching together separate vector, caching, and document stores for every agent was the single biggest drag on their production timelines. Agent Memory gives them a unified, framework-agnostic persistence layer that operates identically in cloud and self-managed environments from cloud to edge, and runs at the latency their agents actually need. That's what it takes to move from pilot to production-and the vendors who understand this will define the infrastructure category for the next decade of AI."