How Postman runs Agent Mode for 40 million developers on Amazon Bedrock

What happened

Building an AI agent (AI that carries out multi-step tasks rather than answering one question) that works in a demo is a different problem from running one for 40 million developers. Postman and AWS share the architectural patterns behind Agent Mode: controlling tool sprawl, exposing schema-based reads, and treating context as the real bottleneck, plus how it runs on Amazon Bedrock at scale.

Building an AI agent for a demo and operating one for 40 million developers are different engineering problems. Postman set out to build Agent Mode , an AI-native way to work across API (the interface one piece of software uses to talk to another) testing , documentation , discovery, and implementation.

The team expected model quality and prompt design to be the hardest problems. The deeper challenges came from integrating an agent into a mature product with years of interface-driven assumptions, a wide surface area, and specialized concepts.