Database vendor Couchbase today announced the launch of Capella IQ, a new AI-powered tool aimed at enhancing developer productivity when building applications on the Couchbase Capella database-as-a-service (DBaaS) cloud platform.

Couchbase was originally developed as an open-source NoSQL database technology and has grown in recent years to add capabilities that are commonly found in relational database technologies. In 2021, Couchbase, Inc., went public on the NASDAQ and now trades under the symbol BASE. That same year, the company first released its Capella DBaaS platform, which has continued to expand with support on multiple cloud platforms including Google Cloud.

The goal with Couchbase Capella is to provide a database platform for developers that is easier to use and manage. The launch of Capella IQ brings the power of generative AI to the platform to help developers write database code.

“Think of it [Capella IQ] as a co- pilot for developers, using LLM [large language model] foundation models to really enhance the productivity of developers,” Matt Cain, Couchbase president and CEO, told VentureBeat.

The new tool fits into Couchbase’s overall four-pronged AI strategy, according to Cain.

Couchbase’s AI strategy includes driving developer productivity, optimizing AI processing, enabling AI-driven applications anywhere and complementing its technology with partnerships. Cain said that Capella IQ addresses the first pillar around developer productivity.

Cain explained that Capella IQ leverages gen AI models to automate tedious development tasks like generating code snippets, sample data sets and unit tests. He noted that developers can access these capabilities directly within the Capella developer workbench through a conversational interface that is designed to have a low barrier to entry.

“It’s completely aligned with how we’re thinking about our AI strategy, but really focused on helping developers be as productive as possible with Capella and enabling next generation applications,” said Cain.

With Capella IQ, Couchbase is using OpenAI’s models to help with code generation. Cain noted that Couchbase may choose to also work with other LLM providers in the future. 

He also emphasized that there are several capabilities in the Capella platform that help to enable the IQ feature beyond just connecting out to an LLM provider. One such feature is the Index Advisor, an existing built-in capability that is able to analyze data queries and provide users with optimization recommendations to improve database index to accelerate response time.

While Couchbase is now jumping into the gen AI era with Capella IQ, it is still missing at least one critical element needed to help power AI applications: Vector embedding support.

This is an increasingly common feature on existing database platforms, with multiple vendors including DataStax, Google with AlloyDB and MongoDB among others, announcing support in 2023.

Support for vector embedding is very much on Cain’s mind, and it is part of his company’s roadmap for inclusion in the near future. He explained that vector embedding support will be enabled in the future as an extension to the platform.

“Our underlying system is a multi model caching JSON document database that performs both operational and analytical capabilities and then we have architecturally enabled services like full text search,” he said. “With a similar architecture we can approach vector and make that a seamless aspect of our platform, where developers can not only take advantage of those capabilities, but do more with less, with a true enterprise platform.”