Klaviyo (NYSE: KVYO) announced a series of platform updates at its annual customer conference K:BOS in Boston, including opening its software to external AI tools and introducing SQL querying capabilities for marketers.

The company said it is making more than 260 Model Context Protocol (MCP) tools and 490-plus APIs accessible directly from third-party AI systems, including Claude and ChatGPT. The move allows marketers or automated agents to access Klaviyo data and execute tasks without logging into Klaviyo's own interface.

Klaviyo is also introducing SQL inside its Klaviyo Data Platform (KDP), currently in preview. The feature allows users to ask questions in plain language through Composer or MCP, which are then translated into SQL queries run against the brand's own data. The tool can currently query profiles, events, segments, lists, and catalogs, with additional object types planned.

The company rebranded its Marketing Analytics product as Personalization, describing it as a layer of specialized models that separately handle audience targeting, content selection, and send timing across channels.

Klaviyo's Composer assistant, used by more than 138,000 users according to the company, gained the ability to schedule recurring tasks, save custom workflows, and edit segments directly. Composer is also now available on a new Klaviyo mobile app, currently in preview.

The company's Customer Agent, which it says has exchanged more than 1.3 million messages with end customers, will add voice as a new channel. The voice feature is described as handling full calls using a customer's history without transfers or holds. The agent also now supports more than 100 languages with automatic locale detection through Shopify Markets.

Additional features announced include Customer Hub web personalization updates, a Clicks to Bricks location-based targeting tool in preview, and a forthcoming Compliance Hub for regulatory tracking.

According to a statement from co-founder and co-CEO Andrew Bialecki, "An agent is only as useful as three things; whether it understands a brand's data, whether it can reach that understanding from wherever it's running, and whether it can act on it for every individual customer."