Amazon Web Services (AWS) has launched the AWS DevOps Agent and AWS Security Agent, SiliconANGLE reported on Tuesday.

The two agents serve as AI teammates available to DevOps and security teams around the clock. They focus on cutting the time people spend on post-incident troubleshooting and helping teams concentrate on preventive optimization.

The DevOps agent monitors applications 24/7 across cloud and on-premises environments and automatically starts investigating when it detects anomalies. It correlates and analyzes telemetry, code and deployment data to identify the cause.

Madhu Balaji (마두 발라지), a senior specialist solutions architect at AWS, said it was designed to work like an experienced site reliability engineer.

The DevOps agent also analyzes past outage patterns and presents recommendations to prevent recurrence. It previously supported only AWS environments, but has expanded support to Microsoft Azure and on-premises systems through the Model Context Protocol, the company said.

The security agent continuously conducts penetration tests across an entire application portfolio.

Penetration testing simulates cyber attacks to find system vulnerabilities, but costs have been a burden because specialists carry out the work.

The security agent does not stop at finding potential vulnerabilities, and instead attempts real multi-step attacks to validate threats. Ayush Singh (아유쉬 싱), an AWS product manager, said it indexes source code and API specifications to identify business logic flaws that other tools can miss. He said that when vulnerabilities are found, it automatically creates a pull request in a git repository to propose code fixes and can cut remediation time from weeks to hours.

The two agents are currently available in U.S. East and West, Europe (Frankfurt and Ireland) and Asia Pacific (Sydney and Tokyo) regions. They are currently provided for free, but the DevOps agent is set to switch to paid service from April 10. A schedule for monetising the security agent has not yet been disclosed.