Home icon

How MoEngage turned over 400 APIs into agent tools with AgentCore Gateway

Machine Learning Blog



This article describes how MoEngage used Amazon Bedrock AgentCore Gateway to turn 400+ internal REST APIs into permission-scoped agent tools without exposing them publicly or requiring individual MCP server implementations.

  • AgentCore Gateway generates MCP tools directly from OpenAPI specs, eliminating need for 20+ separate MCP servers
  • Cedar policies filter tools per agent before serialization, keeping 400-tool catalogs out of context windows
  • Managed egress over Amazon VPC Lattice provides private connectivity across seven AWS Regions without per-API networking
  • Semantic search capability lets agents discover tools by describing tasks instead of memorizing tool names
  • API onboarding reduced from days to hours with simple OpenAPI spec annotations and CI automation
  • Identity travels as headers, not tool parameters, preventing models from accessing or modifying tenant/user context

By separating concerns between product teams (API ownership) and platform teams (protocol/policy/connectivity), MoEngage saved approximately six months of engineering effort and enabled search-first agent design patterns.



Go to article

The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.

Related articles

Oct 22
2026
AgentCore Gateway now supports Bedrock InvokeModel APIs and model fallback
Dec 8
2025
Streamline AI agent tool interactions: Connect API Gateway to AgentCore Gateway with MCP
Oct 28
2026
Building an Agentic AI Platform on Amazon Bedrock AgentCore
Sep 24
2026
Build a multi-account AI agent with AgentCore Gateway and MCP

The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.