Data Management Should Build Down From the Agent
15m
Models Come and Go. Data Stays.
Agentic data management is quickly becoming one of the hardest problems in enterprise AI. Models can be rented, swapped and replaced each year. The data underneath them is where long-term value lives. Agents also generate far more queries than humans ever did. Machines already drive a growing share of search traffic. The demands on data infrastructure are rising just as fast.
Couchbase Chief Product and Strategy Officer Barry Morris examines that shift with Alan Shimel at KubeCon North America 2026. Morris has led several database companies and held database leadership roles at AWS and Google Cloud.
Systems of Engagement, Turbocharged
Morris traces Couchbase back to its roots in CouchDB and Memcached. That DNA still shows in its focus on low latency, high throughput and reliability. NoSQL was built for systems of engagement rather than back-office systems of record. The agentic era, he says, is that same idea at far greater scale.
Couchbase is also inherently distributed. Clusters scale in and out with load and can replicate across regions. Morris says its autonomous Kubernetes operator helps large customers run stateful workloads on cloud-native infrastructure.
Building Down From the Agent
At its Catalyst conference, Couchbase expanded its AI Data Plane. Morris explains the philosophy behind it. Rather than tweaking a database to suit agents, Couchbase designs for agents first and works backward. The latest updates added deeper integration with Amazon Bedrock.
Couchbase also introduced Converge, which lets applications built for MongoDB or Redis run on Couchbase. The goal is to help large enterprises consolidate databases with little recoding. He adds that database APIs are consolidating, too. SQL and the MongoDB API have become the general-purpose choices for many customers.
Kubernetes Gets a Second Wave
Morris and Shimel agree that Kubernetes is enjoying renewed momentum. Much of the AI stack runs on cloud-native infrastructure. Shared standards like Kubernetes, JSON and MCP help the whole industry move faster. For teams tackling agentic data management, those standards make the path to production far less painful.