Official Source· AWS Artificial Intelligence· AI· Published Sep 29, 2026

AWS demonstrates synthetic monitoring with Nova Act and AgentCore

AWS has published a synthetic-monitoring architecture that uses Amazon Nova Act and Amazon Bedrock AgentCore to run browser-based user journeys on a schedule and alert teams when a journey fails. The post includes sample implementation and deployment paths using Python or AWS CDK.

AWS demonstrates synthetic monitoring with Nova Act and AgentCore

Full Report

On September 28, 2026, AWS published a technical walkthrough for building synthetic monitoring with Amazon Nova Act and Amazon Bedrock AgentCore. The approach simulates user actions to regularly check critical workflows such as logins, purchases, and form submissions. AWS says journey-level checks can reveal problems that infrastructure metrics or API checks alone may not immediately surface.

In the example architecture, Amazon EventBridge Scheduler starts runs at set intervals, AgentCore Runtime hosts the agent, and the AgentCore Browser tool supplies an isolated browser environment. If a journey fails at a step, the agent can send a notification through Amazon SNS. The post gives example schedules ranging from every five minutes to hourly, depending on the workflow’s importance.

Nova Act interacts with the site through natural-language actions and visual input rather than fixed DOM selectors. AWS says this can reduce selector maintenance when website interfaces change. The post also advises teams to check outcomes explicitly at important steps and to build in retries for cases where the agent does not adapt successfully.

AWS provides a complete sample implementation and describes deployment using either a Python deploy.py script or AWS CDK. The walkthrough extends AWS’s work on production UI automation following its December 2025 announcement that Amazon Nova Act was generally available. For teams monitoring e-commerce checkout, bookings, or account workflows, the sample offers a concrete pattern that connects scheduled journey tests with browser execution and failure alerts.

Why This Matters

A service or API can appear healthy while a user-facing workflow is broken. AWS’s example connects journey execution to failure alerts and uses visual interaction to lessen reliance on fixed interface selectors. It also makes clear that teams still need explicit outcome checks and retries; the agent’s ability to act is not, by itself, proof that a journey succeeded.

Evidence

Start with the primary evidence, then review supporting sources and context.

1 items

Primary Evidence

1
  • OfficialAWS Artificial IntelligenceDocumentOriginal

    Implementing synthetic monitoring using Amazon Nova Act

    Sarath Krishnan

    “Learn an agent-driven approach to synthetic monitoring using Amazon Nova Act and Amazon Bedrock AgentCore. The post covers the architecture and patterns for resilient, managed user-journey validation that moves beyond brittle UI scripts, with a complete sample implementation.”

Recommended signals

Recommended by brand, topic, and recent related coverage

No more recommended signals right now

Official statement

Open original source