Skip to content

Demonstration project

DevOps foundation for a containerised web application on AWS

A reference architecture showing how we take a small team from manual deployments to a pipeline-driven, Terraform-managed AWS environment.

Demonstration project

This is a demonstration of our approach to a common problem. It describes a reference architecture and is not an account of a specific client engagement.

Project overview

This demonstration describes the foundation we typically build for a startup running two or three services: a web frontend, an API and a background worker.

Business problem

Deployments are performed manually from a developer laptop. Staging and production were created by hand in the AWS console and no longer match. There is no central view of errors or resource usage.

Technical challenge

Introduce automation without stopping feature work, keep the setup simple enough for a small team to own, and avoid over-engineering for the current scale.

Architecture

  • Separate AWS accounts or VPCs for staging and production, defined in Terraform modules
  • Container images built in CI and stored in Amazon ECR
  • Services running on ECS Fargate (or EKS where several services justify Kubernetes)
  • PostgreSQL on Amazon RDS with automated backups
  • Secrets in AWS SSM Parameter Store
  • Prometheus and Grafana for metrics and alerting

Implementation

  • Terraform modules for networking, compute, database and IAM, with remote state and locking
  • CI pipeline: lint, unit tests, image build and vulnerability scan on every merge request
  • Automatic deployment to staging; manual approval gate for production
  • Dashboards for request rate, errors, latency and resource usage, with a small set of actionable alerts
  • Runbooks for deployment, rollback, scaling and restoring from backup

Technology stack

  • AWS
  • Terraform
  • Docker
  • GitHub Actions
  • Prometheus
  • Grafana
  • PostgreSQL

What this approach delivers

  • Every environment can be recreated from code
  • Deployments follow one reviewed, repeatable path
  • The team has documentation for routine operations

← All work

Want this approach applied to your setup?

Tell us what you are working on. We will reply within one business day and suggest a sensible next step.