AIToday
Large Language ModelsAI Coding AssistantsAI Business & IndustryAmazon AI BlogPublished: Oct 10, 2026, 01:00 JST

Postman runs Agent Mode for 40 million developers on Amazon Bedrock

Postman runs Agent Mode for 40 million developers on Amazon Bedrock

3 Key Points

  1. What happened

    Postman and AWS described the architecture behind Agent Mode, which runs on Amazon Bedrock and serves 40 million developers. In Postman's testing, tool-selection errors increased once the visible toolset exceeded approximately 40 tools.

  2. Why it matters

    The team said missing or incomplete context caused more failures than missing capabilities, so production agents appear to hinge on managing what the model sees, not just adding more tools.

WHO IT HITSEnterprise IT and platform teams running customer-facing AI agents, along with product managers at mature software companies, can use Postman's patterns to plan tool scoping, data access, and context budgets when moving agents from demo to production.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, in your inbox every morning.

Context & Analysis

Postman's Agent Mode is a portal for working across API testing, documentation, discovery, and implementation in an AI-native way. Postman has evolved over 11 years, and developers and users learned to locate information by expanding sidebars, checking tabs, and opening requests. The team built dedicated context handlers that distilled each entity into what the agent needed to know, because serializing the existing interface data model did not produce useful context.

Two other patterns came out of that work. The architecture selects tools based on need and context and isolates individual execution threads, so the model sees only the tools relevant to the current task. For products such as the API Catalog, Postman consolidated multiple narrow views into a single query tool that can generate complex queries with joins and WHERE clauses, shifting the engineering job from building one tool per question to modeling the data well once.

On Amazon Bedrock, Postman gains model flexibility across supported Anthropic Claude models, cross-Region inference for throughput, model-dependent zero data retention, and multi-tier prompt caching. Postman's production implementation is proprietary and is not available as a public sample repository.

FAQ
How many tools does Agent Mode show the model at once?
A root agent queries a vector database of tool embeddings and narrows more than 170 tools to approximately 15 relevant to the request, then hands those to a context-isolated sub-agent.
How does Postman keep enterprise data private in Agent Mode?
Postman uses Amazon Bedrock Guardrails to redact personally identifiable information before it reaches the underlying large language model, and enterprise admins can turn this on in Agent Mode's guardrail settings.
Does Agent Mode use prompt caching to cut costs?
Yes. Agent Mode uses Amazon Bedrock prompt caching, with a one-hour checkpoint for the near-immutable core and a five-minute checkpoint for more variable context.
Amazon AI BlogRead Original Article

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Questions and answers are published on this page.

Related Articles

Next articleClaude Code traces highlight 26-year-old suspect in South Korea attacks