en.Wedoany.com Reported - Couchbase has launched a unified data infrastructure layer for enterprise AI agents, called the AI Data Plane, designed to provide persistent memory, contextual retrieval, and an operational layer for data spanning from the cloud to the edge.
Couchbase states that by bridging the gap between fragmented data services, its product helps enterprise agents make consistent decisions at scale. Barry Morris, Chief Product and Strategy Officer at Couchbase, noted that the database layer is the key to scaling or stalling agentic AI, yet most in the industry still treat agent memory as an afterthought. The company built the AI Data Plane because customer feedback indicated that stitching together separate vector, cache, and document stores for each agent was the biggest obstacle on their production timelines.
Data platform vendors often point out that enterprise customers have undiscovered resources hidden across fragmented pools of structured and unstructured data. Couchbase's AI Data Plane aims to alleviate the cumbersome management of retrieving relevant information from these sources by establishing a single distributed platform that converges vector, document, cache, and operational data. Competition in this space is intensifying, with Couchbase joining vendors such as Databricks' Unity Catalog, Snowflake's Horizon platform, and Everpure (formerly Pure Storage) with its data intelligence offerings.
Couchbase says its product integrates Agent Memory, a tool catalog for discoverable agent tools, and an enterprise-supported, self-managed Model Context Protocol (MCP) server that can run on its Capella and self-managed environments. The AI Data Plane also leverages the vendor's distributed multi-model architecture, which supports JSON documents, key-value pairs, and SQL for JSON queries within a single system.
Barry Morris added that Agent Memory provides users with a unified, framework-agnostic persistence layer that operates consistently from cloud to edge in both cloud and self-managed environments, and at the latency agents actually require. He believes this is what is needed to move from pilot to production. Conversational AI platform Agora is one of the early adopters of Couchbase's product. Patrick Ferriter, Senior Vice President of Product at Agora, noted that every one of its conversational AI use cases requires efficient data retrieval to feed the AI agent pipeline, whether for outbound sales, customer service, physical AI, or entirely new domains. The company has a multi-year partnership with Couchbase, and this is a natural extension of that collaboration as it expands into agentic workloads.
The launch of the AI Data Plane is Couchbase's largest product update since it was acquired and taken private by Haveli Investments in a $1.5 billion all-cash transaction last September.
