Klaviyo has moved its AI marketing agent, Composer, into public beta and expanded its Customer Agent platform, both built on the same real-time customer profile. The launch connects marketing and service teams inside a single CRM, tackling a persistent problem: fragmented data that forces customers to repeat themselves and causes brands to miss revenue opportunities.
The two agents share intelligence across marketing and customer service. When Customer Agent resolves a support interaction, it writes things like product interests, preferences and intent signals back to the customer record. Composer can then build more targeted campaigns based on those signals. Engagement data from marketing campaigns, in turn, tells Customer Agent how to personalize future service interactions.
"Businesses aren't struggling because they lack AI tools-they're struggling because most AI can't act on the context that matters," said Jamie Domenici, Klaviyo's Chief Marketing Officer. "The next era of consumer experiences will be defined by companies that can combine AI with deep customer understanding to move faster, personalize better, and turn customer insights into growth."
The risk of disconnected agents is not theoretical. Paul Adams, Chief Product Officer at Fin (formerly Intercom), described what happens when separate AI tools operate with different memory. "If you have multiple agents-one for sales, one for service, one for success-especially if these agents come from different companies… They don't share the same context. They have different customer history, different memory. And it'll end up being a big mess."
Shared customer profile drives agent coordination
Klaviyo's approach is to anchor both agents in a single customer record. This shared-context model directly addresses the handoff points between teams and tools that can frustrate customers. Instead of treating service interactions as separate from revenue, support conversations become a commercial intelligence source. A resolved service ticket might reveal purchase intent, churn risk or product dissatisfaction-signals that improve marketing campaigns and customer journeys.
Composer enters public beta
Composer is designed to help marketing teams find revenue opportunities and build campaigns faster. Users can ask the agent to audit live campaigns, flows and segments, and it surfaces ranked opportunities: an underperforming abandoned-cart flow, a welcome journey where customers drop off, or a lapsed segment of high-value buyers. Once a marketer selects an opportunity, Composer recommends actions and builds the campaign-including audience, content and channel execution for email and SMS. Klaviyo says the agent draws on 14 years of customer context and patterns from nearly 200,000 brands, plus each brand's own performance data. Campaigns are prepared for review and launch; nothing goes live without approval.
Customer Agent adds service actions with full context
Customer Agent uses the same customer data that powers Composer. Many AI service tools lack visibility into what a customer bought, what is in their cart or which marketing messages they've received. Klaviyo argues that this full profile lets Customer Agent deliver more relevant support. The agent can now be configured to create new service experiences with a simple prompt-for instance, a retailer can describe a product-finding quiz and Klaviyo AI generates the questions, logic and product recommendations.
The agent can also take action through connectors and APIs, completing returns and applying loyalty points across web chat, email, SMS, WhatsApp and custom surfaces. This capacity fits a broader shift in AI service design. As Dani Wanderer, Chief Marketing Officer at Ada, put it: "Leaders invest in system integration. They give their AI agents both read and write access to the backend systems. So whether it's CRM or order management, billing systems, that's what enables the AI to do things for consumers, not just answer questions."
Why this matters for Customer Support
For support professionals, this architecture changes how service work creates value. Instead of resolving a query and closing the ticket, agents can capture revenue-critical signals and feed them back into the customer record automatically. A support interaction that reveals buying intent or product dissatisfaction can trigger a follow-up campaign or adjust the customer's profile for future service. The result is a system where service directly influences revenue, and AI reduces the manual handoffs that often lead to dropped context and repeat contacts.
For those who want to build practical skills in this area, an AI Learning Path for User Support Specialists covers the automation and integration patterns that make these kinds of feedback loops work. Composer public beta is available now. Customer Agent is available as a full AI-powered service platform, with additional features scheduled throughout 2026.
Your membership also unlocks:
