How I Built an AI Agent That Cut My AWS Bill by 40% (CrewAI + Bedrock)

Описание и применение

Автоматический агент CrewAI + Bedrock сканирует AWS‑аккаунт, выявляет неиспользуемые ресурсы и генерирует отчёт с точными экономическими выгодами, сократив счёт на 40 %.

Задача

AWS accounts often accumulate hidden costs from orphaned resources, idle IPs, and outdated storage, making manual audits time‑consuming and error‑prone.

Результат

The system identified $125/month in waste, including orphaned EBS volumes, unused Elastic IPs, and a non‑reserved EC2 instance, enabling quick actions that saved $40/month before lunch.

Рабочий процесс

• Clone the repo and install dependencies
• Set AWS_DEFAULT_REGION environment variable
• Run main.py to execute the three agents sequentially
• Review the generated markdown report
• Act on quick‑win recommendations (delete orphaned volumes, release unused IPs, reserve instances)

Доказательства

It found $125/month I was burning for nothing.

Источник

Three orphaned EBS volumes: $33/month. Two unattached Elastic IPs: $7.20/month.

Источник