AWS Systems Manager (SSM) is an AWS managed service for centrally operating and managing EC2 instances and on-premises servers. It enables operational tasks such as patch application, command execution, parameter management, and inventory collection to be performed in a unified manner, without the need to connect to each server individually via SSH or RDP.
When you only have a handful of servers, logging in via SSH to perform tasks is sufficient—but that changes as the number grows. Applying the same patch to dozens of machines, keeping track of installed packages across all of them, collecting logs in bulk during an incident: handling these tasks manually is simply not practical. SSM provides a mechanism for executing these operational tasks in bulk from the AWS console or CLI.
SSM is not a single service but is composed of multiple feature groups. The most notable are described below.
Run Command is a feature that remotely executes shell scripts and PowerShell commands on managed nodes. Because it requires no open SSH ports and allows execution permissions to be controlled via IAM, it is easier to manage than traditional SSH connections from both a security and operational standpoint. No additional charges apply.
Parameter Store is a feature for securely storing and distributing configuration values such as database connection strings and API keys. It supports encryption via KMS, and the standard pattern is for applications to retrieve values using aws ssm get-parameter.
Patch Manager scans the patch compliance status of operating systems and automatically applies patches based on a defined baseline. Combined with Maintenance Windows, it enables a workflow where patch application is completed outside of business hours.
Session Manager provides browser-based shell access. It eliminates the need for bastion servers, and the fact that session activity logs are automatically recorded in CloudTrail and S3 makes it particularly valuable in environments with strict audit requirements.
Each SSM feature operates through the SSM Agent installed on managed nodes. Because it comes pre-installed on AMIs such as Amazon Linux 2 and Windows Server, EC2 instances can often be used without any additional setup. On-premises servers and edge devices require a manual installation, but by following a registration procedure called hybrid activation, they can be added as managed targets in the same way as EC2 instances.
In the author's experience, version discrepancies in SSM Agent have occasionally been the root cause of issues. When Run Command fails on only specific nodes, the standard first step is to suspect the Agent version.
Run Command, Session Manager, and Parameter Store (Standard parameters) are available at no additional charge. Fees do apply for certain advanced features, such as Parameter Store Advanced parameters and OpsItem operations in OpsCenter. The ability to keep basic operational management costs low even in large-scale environments is one of SSM's strengths.



A multi-agent system is an architecture in which multiple AI agents divide roles and coordinate with each other to accomplish a shared objective.

Firecracker is an open-source virtual machine monitor (VMM) developed by AWS that boots lightweight microVMs in under 125 milliseconds, achieving both container-level density and VM-level security isolation.

MLOps is a practice that automates and standardizes the entire lifecycle of machine learning model development, training, deployment, and monitoring, enabling the continuous operation of models in production environments.

Shadow AI refers to the collective term for AI tools and services used by employees in their work without the approval of the company's IT department or management. It carries risks of information leakage and compliance violations.

SLM (Small Language Model) is a general term for language models with a parameter count limited to approximately a few billion to ten billion, characterized by the ability to perform inference and fine-tuning with fewer computational resources compared to LLMs.