MoEngage shares agent architecture based on AgentCore
Understand this story
MoEngage has disclosed how its Merlin AI agent system is built: based on Amazon Bedrock and Amazon Bedrock AgentCore, it lets marketing teams complete campaign drafts, analysis, and adjustments through conversation. MoEngage says that for its customer Glance, push notification click-through rates increased by 38% as a result. MoEngage pays per token rather than building its own GPU infrastructure. Amazon Bedrock does not use customers' prompts or responses to train foundation models. Feature availability of AgentCore varies by AWS Region.
AI overview based on source reports · Updated 17 min ago
Coverage timeline
Follow the coverage to explore different perspectives.
- AWS Machine Learning BlogHow MoEngage built agentic customer engagement on Amazon Bedrock AgentCore
MoEngage built Merlin AI, a set of agents running on Amazon Bedrock AgentCore runtime that give marketing and product teams conversational access to campaigns, analytics, segments, content and flows, using Anthropic Claude models on Amazon Bedrock for reasoning and generation. MoEngage exposed its existing OpenAPI 3.0-compatible platform APIs as MCP tools via the FastMCP framework, and kept tenants isolated by reusing JWT authentication, passing workspace and user context in headers rather than tool arguments, and running each user session in its own isolated AgentCore runtime environment. MoEngage reports that customer Glance reached campaign go-live 50% faster with a 38% uplift in push click-through rates, and that managed pay-per-token inference reduced AI infrastructure costs by more than 50% versus self-hosted models.
Activity for this story
Not enough continuous observations to show a trend yet.