MoEngage Merlin AI Introduces Agentic Customer Engagement Built on Amazon Bedrock AgentCore
MoEngage has launched Merlin AI, a set of agent-powered customer engagement tools running on Amazon Bedrock AgentCore. Merlin AI turns existing platform APIs into agent-callable MCP tools, secures multi-tenant data, and supports conversational campaign management and analytics using Claude Sonnet models. This update streamlines campaign workflows and provides robust access control.
What changed?
MoEngage has released Merlin AI, a suite of agentic customer engagement capabilities built on Amazon Bedrock and AgentCore. The company converted its platform APIs, which are OpenAPI 3.0-compatible, into agent-callable Model Context Protocol (MCP) tools using the FastMCP framework. Merlin AI includes four types of agents—Embedded, Conversational, Custom, and Decisioning—that allow users to draft, analyze, and adjust campaigns via a conversational interface. All operations happen securely with strict tenant isolation, using JWT authentication handled at the AgentCore runtime level. Agents can now access live platform data and leverage Anthropic Claude Sonnet models for reasoning and execution.

Why does it matter to an everyday developer?
Developers integrating or extending their customer engagement solutions benefit from a model where APIs become reusable MCP tools, enabling natural language interfaces and agent automations without rebuilding backend logic. The FastMCP-based approach streamlines tool creation as API specifications evolve. Robust session isolation and tenant data controls built into AgentCore runtime reduce risks and complexity of securing sensitive customer data. Merlin AI's use of Amazon Bedrock models ensures teams do not need to manage model lifecycle, infrastructure, or model-specific security. Developers can offer end-users a unified, conversational experience for campaign management, analytics, and automation without exposing underlying integration and authentication complexity.
What can the developer do now?
Developers working with MoEngage can tap into Merlin AI agents to automate and streamline campaign-related workflows. By defining or updating OpenAPI 3.0 specs for their APIs, they can automatically generate new MCP tools with FastMCP, which become immediately available to agents for natural language interaction or decisioning. Developers can take advantage of Merlin AI's conversation-based interfaces to eliminate siloed workflows for marketing users, integrate live analytics and content management, and configure tenant-specific access controls. No major action is required for existing MoEngage users, but teams can review and optimize their API specs to maximize future agentic integrations.
