How I Use AI to Design and Implement Cloud Infrastructure with Terraform

The Problem

Migrating to the cloud is hard. Doing it fast without breaking production is even harder. Most teams either:

• Write Terraform by hand, making mistakes along the way
• Use blueprints that don’t fit their specific needs
• Skip security hardening to save time

The AI Solution

I use AI as my “junior DevOps engineer” - it generates initial infrastructure code that I then harden and refine. Here’s how it works.

The Workflow

1. Assessment    → Understand requirements, existing setup
2. Design        → Plan VPC, subnets, security
3. AI Generation → Prompt AI for Terraform code
4. Manual Refinement → Apply security hardening
5. Testing       → Plan/validate in non-prod
6. Implementation → Apply to production

AI in Action: Example Prompts

Prompt 1: VPC Setup

"Create Terraform for AWS VPC with:
- 10.0.0.0/16 CIDR block
- 2 public subnets in different AZs
- 2 private subnets in different AZs
- Internet gateway
- Proper route tables"

AI Output: Complete VPC, subnets, IGW, route tables - about 80% done.

Manual Refinement Needed:

• Add proper tags for cost tracking
• Enable DNS hostnames/support

Prompt 2: Security Groups

"Create security group for Flask API allowing HTTP, HTTPS, SSH from VPC only"

AI Output: Basic rules but allows SSH from anywhere.

Manual Refinement:

# Restrict SSH to admin network only
ingress {
  from_port   = 22
  to_port     = 22
  protocol    = "tcp"
  cidr_blocks = ["10.0.0.0/16"]  # VPC only, not 0.0.0.0/0
}

Prompt 3: Database

"Add RDS PostgreSQL with private subnets, encryption, daily backups"

AI Output: Basic database definition.

Manual Refinement:

• Enable storage_encrypted = true
• Set backup_retention_period = 7
• Add to DB subnet group

Code Example: AI vs Manual

From POC #1: Terraform API:

# AI-generated (base structure)
resource "aws_instance" "api" {
  ami           = data.aws_ami.ubuntu.id
  instance_type = "t3.small"
  subnet_id     = aws_subnet.public[0].id
}

# After manual hardening
resource "aws_instance" "api" {
  ami           = data.aws_ami.ubuntu.id
  instance_type = var.instance_type
  subnet_id     = aws_subnet.public[0].id
  
  vpc_security_group_ids = [aws_security_group.api.id]
  iam_instance_profile   = aws_iam_instance_profile.api.name
  
  root_block_device {
    encrypted = true
  }
  
  user_data = <<-EOF
    #!/bin/bash
    apt-get update && apt-get install -y python3-pip nginx
    # ... app setup
  EOF
}

Results

What This Means for Your Organization

• Faster Time-to-Market: AI generates infrastructure in minutes, not days
• Better Security: Manual hardening ensures production-grade security
• Lower Costs: Tags enable accurate cost allocation
• Reproducible: Code can be version-controlled and reviewed

Ready to accelerate your cloud migration? Contact me for a free 30-minute assessment.