What changed

MoEngage, a customer engagement platform used by over 1,350 brands, has integrated AI agents powered by Amazon Bedrock to automate and enhance customer engagement workflows. These agents, part of a system called Merlin AI, allow marketing and product teams to draft, analyze, and adjust campaigns from a single conversation, with agents deciding what to send each customer. The integration leverages Amazon Bedrock AgentCore to provide secure, tenant-scoped access to live data without rebuilding existing APIs.

What this means in practice

Marketing teams can now use AI agents to perform tasks such as generating personalized copy, creating marketing banners, and building customer journeys. For example, a team member can request a campaign for users who abandoned a cart, and the agent will handle the segmentation and content creation. The system also supports conversational agents that answer questions or act on user inputs across campaigns, analytics, and segments. Custom agents can run continuously on MoEngage data, with options for full autonomy or human review. Decisioning agents handle individual customer interactions, selecting the best offers and delivering personalized messages across channels.

Limitations and availability

The integration is available to MoEngage customers through the MCP Connector, which provides secure access to the platform’s APIs. The solution uses foundation models like Anthropic Claude Sonnet 4.6, 5, and 5.5 on Amazon Bedrock for reasoning, tool-call planning, and content generation. The architecture ensures data isolation for each tenant, with the MCP server hosted on AgentCore runtime. The system relies on existing OpenAPI 3.0-compatible APIs, which enforce permissions and data residency rules. The supplied announcement does not specify exact performance benchmarks or availability timelines for the full system.

Source facts for MoEngage integrates AI agents with Amazon Bedrock for customer engagement: A marketing or product team member planning a weekend flash sale often opens three tools before writing a single line of copy. Repeat that across email, push, SMS, and WhatsApp, and much of a marketing team’s week goes to navigating tools instead of making decisions. This capability points to a broader shift in software platforms.
Original AINewsTrend explainer based on the cited source. This graphic summarizes source statements; it is not an independent benchmark.

Sources and further reading

  1. awsml: How MoEngage built agentic customer engagement on Amazon Bedrock AgentCore