MoEngage has introduced Merlin AI Custom Agents, a set of marketer-configurable AI agents designed to automate lifecycle and CRM workflows with visibility, permissions, guardrails, and Model Context Protocol connectivity.
MoEngage has launched Merlin AI Custom Agents, a product aimed at helping lifecycle marketing and CRM teams build governed AI agents for marketing workflows.
According to MoEngage’s announcement distributed via PR Newswire, the new capability is part of the company’s Merlin AI portfolio and is designed to let marketers define custom agents with visibility into actions, marketer-set guardrails, and connectivity to external AI systems through an open Model Context Protocol, or MCP, architecture.
MoEngage says Merlin AI Custom Agents are intended for teams that want to automate parts of customer engagement work while retaining control over how AI systems operate. The PR Newswire announcement describes the agents as tools for lifecycle and CRM teams to build marketing workflow agents that can be governed by business users rather than left as opaque automation.
The company’s own “What’s New in MoEngage?” product update page lists Merlin AI Custom Agents alongside Merlin AI native agents and the MoEngage AI Connector. MoEngage describes these Spring 2026 updates as part of an “open” and “explainable” AI direction for automating multi-channel growth workflows.
MoEngage positions the launch around three themes: giving marketers visibility into agent activity, allowing marketers to define guardrails, and connecting MoEngage to external AI tools through MCP. Those points are important because many marketing teams are experimenting with AI assistants but remain concerned about brand safety, approvals, data permissions, and accountability.
A separate MoEngage documentation page explains that the company provides a hosted MCP server and connector for linking AI assistants to MoEngage workspaces. In that documentation, MoEngage says the MCP server can connect tools such as Claude and ChatGPT to MoEngage for tasks including campaign search, analytics, and content review.
MoEngage’s documentation also says the MCP setup uses OAuth authentication and role-based permissions. Those controls matter in a marketing context because customer engagement platforms can include campaign data, audience segments, performance analytics, and other operational information that should not be exposed broadly.
The MCP reference is also notable because Model Context Protocol has become a common way for AI applications to connect to external tools and data sources. MoEngage is not presenting the feature only as an in-product assistant; its documentation indicates that external AI systems can interact with MoEngage through a standardized connection layer, subject to authentication and permissions.
MoEngage’s announcement emphasizes “full visibility” and “marketer-defined guardrails.” Based on the company’s description, the product is aimed at users who want to delegate repetitive or complex workflow steps to AI while keeping boundaries around what those systems can do.
The available source material does not provide a full technical breakdown of every guardrail or approval mechanism. However, MoEngage’s PR Newswire announcement and product materials consistently frame the release as a governed approach to AI automation for customer engagement teams.
That framing reflects a broader pattern in enterprise AI product development: vendors are moving beyond general chat interfaces toward role-specific AI systems that can take action inside business software. For marketing platforms, that can include finding campaign information, reviewing content, analyzing performance, or supporting multi-channel engagement workflows. In MoEngage’s case, the company’s documentation specifically mentions campaign search, analytics, and content review as MCP-connected capabilities.
Lifecycle and CRM teams often manage many recurring tasks across email, push notifications, in-app messaging, SMS, and other customer channels. MoEngage’s product update page describes its recent AI releases as supporting automation for multi-channel growth, which suggests the company is targeting teams that need to coordinate campaigns across several touchpoints.
The central question for customers will be how much practical control Merlin AI Custom Agents provide in day-to-day use. MoEngage’s materials point to visibility, permissions, OAuth authentication, role-based access, and marketer-defined guardrails, but organizations evaluating the product will still need to review how those controls map to their compliance requirements, approval processes, and internal data policies.
For now, the launch shows MoEngage joining the enterprise software push toward configurable AI agents rather than single-purpose AI assistants. The company’s distinguishing claim is that these agents are designed for marketing workflows while remaining connected to external AI systems through an open MCP architecture.
MoEngage has launched Merlin AI Custom Agents, a product aimed at helping lifecycle marketing and CRM teams build governed AI agents for marketing workflows.
What MoEngage is introducing MoEngage says Merlin AI Custom Agents are intended for teams that want to automate parts of customer engagement work while retaining control over how AI systems operate.
The PR Newswire announcement describes the agents as tools for lifecycle and CRM teams to build marketing workflow agents that can be governed by business users rather than left as opaque automation.
Continue reading
