Why I migrated from legacy hosting to AWS CloudFormation (and the 30% cost savings that followed)

I led the infrastructure migration for Easley Finance, moving from a manually managed Docker container on a traditional host to a fully automated AWS CloudFormation stack. The results: a 30% reduction in hosting costs, zero-downtime deployments, and 99.9% uptime.

The legacy problem

The startup relied on a single Docker container, manually deployed via SSH. Every update required manual restarts. Scaling meant spinning up new servers by hand. Database backups were unreliable at best.

The CloudFormation solution

I designed a multi-environment architecture using CloudFormation templates, enabling the team to spin up identical infrastructure for development, staging, and production with a single command.

The stack runs on AWS Fargate for compute (no servers to manage) and Amazon RDS for the database, with automated backups, health checks, and scaling policies baked into the template.

The 30% cost savings came primarily from right-sizing: the old provider charged flat rates for oversized instances, while Fargate bills only for actual usage. Adding scheduled scaling for off-peak hours cut costs further.

Repository

github.com/hgosansn/fargate-rds-stack