100% Free & Open Source

AI That Investigates
Your AWS Alarms.

Open-source Terraform modules that put AI agents in your AWS account. An alarm fires. The agent checks metrics, logs, and configs. Root cause lands in Slack in seconds. No SaaS. No API keys. Your data stays yours.

AWS Investigation
6 AI Tools
Least Privilege
Read-Only
Per Alarm
~$0.001
main.tf 
module "cloudwatch_ai_alerts" {
  source = "github.com/AIOpsCrew/cloudwatch-alarm-agent//terraform/modules/cloudwatch-slack-alert"

  email             = "you@company.com"
  slack_webhook_url = var.slack_webhook_url
  function_name     = "cloudwatch-ai-agent"
}
Open Source

Production-Ready Terraform Modules

Put AI-powered tools in your own AWS account. No vendor lock-in. No subscriptions. Just terraform apply.

CloudWatch AI Agent

Free • Open Source

This module adds AI alarm investigation to your AWS account. When a CloudWatch alarm fires, an AI agent checks your metrics, logs, and resource configs. It finds the root cause and sends you a Slack message with the fix. All powered by Amazon Bedrock. All inside your account.

AI Investigation
6 Tools
Per Alarm
~$0.001
Response Time
Seconds
  • AI agent checks alarms using 6 read-only AWS tools
  • Reads CloudWatch metrics, EC2/RDS/Lambda configs, logs, and alarm history
  • Sends Slack alerts with root cause and ready-to-run CLI commands
  • Runs fully in your AWS account. No data leaves your environment.
  • Uses Amazon Bedrock (Nova Lite). Bring your own model if you prefer.
  • Set up in under 5 minutes with one terraform apply

n8n Fargate Cluster

Free • Open Source

Run n8n workflow automation on AWS Fargate. This module sets up a high-availability cluster with Aurora Serverless v2 and zero-downtime deploys.

High Availability
Multi-AZ
Auto-Scaling DB
Aurora v2
AWS Cost
~$160/mo
  • Runs across multiple AZs with load balancing built in
  • Aurora Serverless v2 scales up and down based on demand
  • Secure by default: private subnets, least privilege IAM, encryption
  • Zero-downtime deploys for smooth n8n upgrades

Jenkins Sentinel

Private Beta

AI finds why your Jenkins builds fail. It reads the logs, sorts failures into 5 types, and sends a Slack message with the fix. All using AWS Bedrock.

AWS Bedrock
Any Model
Failure Types
5 Categories
AWS Cost
~$22/mo
  • Sorts failures into 5 types: Env, Security, Code, Deps, Config
  • Strips secrets from logs before sending them to the LLM
  • Slack alerts include root cause and confidence score
  • Serverless and event-driven with SQS for reliability
How It Works

From Alarm to Root Cause in Seconds

A fully automated pipeline that runs in your AWS account.

  1. Alarm Triggers

    A CloudWatch alarm crosses its threshold. It sends a message to SNS.

  2. AI Investigates

    The Bedrock agent reads metrics, resource configs, logs, and alarm history.

  3. Root Cause Found

    The AI reviews the data and writes clear steps to fix the problem.

  4. Slack Alert

    Your team gets a rich message with the root cause and CLI commands to run.

Set Up AI Monitoring in 5 Minutes

Open source. Runs in your account. No API keys. No subscriptions. No vendor lock-in.

Contact

Get in Touch

Questions, feedback, or just want to say hey.

Newsletter

New modules, updates, and tips

Subscribe

Issues & PRs

Report bugs or contribute code

GitHub