Cloud Agents (AWS)
Deploy AI Agents to the Cloud

Two paths to cloud AI agents. EC2 for dedicated instances with native install. ECS Fargate for serverless containers. Both managed by Terraform, deployed with one command.

Works great for:

🚀
Startup Scale to cloud
👥
Team Distributed agents
👑
AI-First Org Heavy compute

Offload AI workloads to dedicated cloud infrastructure

Two Paths to Cloud

Choose the right deployment model for your workload

EC2

Dedicated Instance

A full VM running your AI agent natively. Node.js, tmux, and AI CLIs installed directly on ARM64 Graviton hardware. Nginx + SSL included.

Native install (no Docker overhead)
HTTPS with auto-renewing SSL certificates
ARM64 Graviton (cost-effective)
SSH access for debugging
BEST FOR

Long-running agents, persistent workloads, SSH access needed

ECS Fargate

Serverless Container

Serverless containers on AWS Fargate. AI Maestro auto-builds your Docker image, pushes to ECR, and runs it. No servers to manage.

Auto-builds Docker image from Dockerfile
No servers to patch or maintain
Pay only while running
Application Load Balancer included
BEST FOR

Burst workloads, cost optimization, zero-maintenance infrastructure

How It Works

Deploy cloud agents from the dashboard or CLI

1

Prerequisites

Install Terraform and configure AWS credentials:

# Install Terraform
brew install terraform

# Configure AWS
aws configure
2

Create from Dashboard

Open the agent creation wizard, pick EC2 or ECS, fill in the required fields, and deploy:

EC2 REQUIRES:
Domain, SSL Email, SSH Key Name
ECS REQUIRES:
Nothing (all auto-configured)
3

Or Use the CLI

Deploy directly from the command line:

# EC2 with SSL
aimaestro-agent.sh create my-api \
  --ec2 --domain api.example.com \
  --ssl-email admin@example.com \
  --key-name my-key

# ECS Fargate (auto-builds image)
aimaestro-agent.sh create worker --ecs
4

Manage from Dashboard

Cloud agents appear in the AI Maestro dashboard alongside local agents. Same terminal access, same messaging, same management. Destroy infrastructure when done.

What Gets Created

Terraform manages all infrastructure automatically

EC2 Infrastructure

  • VPC with public subnet
  • EC2 instance (Graviton ARM64)
  • Security group (ports 22, 80, 443)
  • Elastic IP address
  • Nginx reverse proxy with SSL
  • systemd service for agent-server
  • Node.js 20, tmux, AI CLIs

ECS Infrastructure

  • VPC with public/private subnets
  • ECS Fargate cluster + service
  • ECR repository (auto-created)
  • Application Load Balancer
  • IAM roles + task execution role
  • CloudWatch log group
  • Docker image (auto-built + pushed)

Perfect For

Heavy Compute Workloads

Your laptop fan is screaming. Move resource-intensive agents to cloud instances with dedicated CPU and memory. Keep your local machine for what it does best.

🌐 Always-On Agents

Agents that need to run 24/7 monitoring builds, processing data, or responding to messages. EC2 instances keep working when your laptop sleeps.

📈 Burst Scaling

Need 10 agents for a sprint, then 2 for maintenance? ECS Fargate scales to zero when idle. Pay only for what you use.

🔒 Isolated Environments

Client work that requires network isolation, compliance boundaries, or dedicated infrastructure. Each cloud agent runs in its own VPC.

Deploy to AWS
With One Command

From your terminal to the cloud in minutes. Terraform handles the infrastructure, AI Maestro handles the agents.

Other Setup Modes