MoEngage built agentic customer engagement on Amazon Bedrock AgentCore

Executive Briefing
“MoEngage developed its Merlin AI assistant system by converting existing platform APIs into...”
Key Takeaways
- 1.MoEngage uses foundation models on Amazon Bedrock including Anthropic Claude Sonnet versions for reasoning and content generation.
- 2.MoEngage customer Glance reached campaign go-live 50 percent faster and saw a 38 percent uplift in push notification click-through rates.
- 3.The architecture uses JSON Web Token validation and passes user context in request headers to keep identity out of the model's reach.
MoEngage developed its Merlin AI assistant system by converting existing platform APIs into Model Context Protocol tools hosted on Amazon Bedrock and AgentCore runtime.
Key Developments and Technical Context
MoEngage uses foundation models on Amazon Bedrock including Anthropic Claude Sonnet versions for reasoning and content generation.
MoEngage customer Glance reached campaign go-live 50 percent faster and saw a 38 percent uplift in push notification click-through rates.
The architecture uses JSON Web Token validation and passes user context in request headers to keep identity out of the model's reach.
Strategic Analysis and Industry Implications
Automating the conversion of OpenAPI specs into Model Context Protocol tools helps enterprise software platforms connect legacy systems to AI agents without rewriting core codebases.
Outlook and Key Factors to Watch
The source text does not specify detailed operational costs or potential latency challenges when scaling multi-layered agent systems across thousands of enterprise consumer brands.







