AI Agent That Investigates Cloudwatch Alarms and Delivers Root Cause Analysis

I built this because I was tired of the manual investigation loop every time a CloudWatch alarm fired: check metrics, grep logs, run CLI commands, piece it together. Takes 15-30 minutes minimum per alarm.

CloudWatch AI Agent automates this: when an alarm triggers, an AI agent autonomously queries your AWS environment (read-only access), analyzes the data, and delivers root cause analysis with actionable AWS CLI commands to Slack.

How it works: Deploys via Terraform module (Apache 2.0 licensed) Lambda function gets triggered by SNS when alarm fires AI agent uses read-only tools to query CloudWatch metrics, logs, EC2/RDS/Lambda configs, alarm history Performs analysis and generates recommendations Sends rich Slack notification with findings and ready-to-run commands

I'm a solo dev/AWS engineer who built this for my own use and decided to productize it. Would love feedback on the approach, pricing model, or technical architecture.

Site: https://aiopscrew.com

Happy to answer any questions about the implementation or design decisions.