AWS DevOps Agent and Datadog MCP Server have finally reached production-ready status for autonomous incident resolution. This collaboration is the result of years of development, first showcased in a December 2025 demonstration.
**Autonomous Incident Resolution: A New Era in DevOps**
The joint effort between AWS and Datadog enables AI-powered agents to automatically identify and resolve system issues without human intervention. This is achieved by combining AWS DevOps Agent, a container-centric deployment tool, with Datadog MCP Server, a monitoring and analytics platform designed for complex systems.
The key to this system lies in the use of machine learning algorithms to analyze data from various sources. By processing this data in real-time, the system can detect anomalies and potential issues before they escalate into full-blown incidents.
**How It Works**
When an issue arises, the AWS DevOps Agent and Datadog MCP Server collaboration spring into action. The agent automatically captures relevant data from the affected system, which is then sent to Datadog’s MCP Server for analysis. This analysis involves predictive modeling and machine learning techniques to determine the root cause of the issue.
The system’s AI engine then generates a plan to resolve the issue, which is executed by the AWS DevOps Agent. This process reduces mean time to resolve (MTTR) and minimizes downtime for systems and applications.
**What This Means**
For organizations relying on cloud-native systems, this new autonomous incident resolution capability is a major breakthrough. With the ability to automatically identify and resolve issues, teams can reduce manual intervention and free up resources for more strategic work. Moreover, this technology can also improve the overall reliability and performance of complex systems.
Bharadwaj Tanikella, AI/ML Product Engineering Leader at Datadog, and Mohammad Jama, Product Marketing Manager, emphasize that this collaboration is a crucial step towards a more autonomous and efficient DevOps ecosystem.
