Published on 2025-07-16 9:20 by Kevin Coyle

Devops let’s goooo

From Events to Enterprise: What changed between “meeting page” and now

Here’s a concise tour of the work that transformed the project from a polished meetings page into a production-ready, cloud-deployable platform with infra-as-code, reverse proxying, and a cleaner developer workflow.

I had the pleasure to work in this cool pattern with a friend/CTO Nash Taylor. He wrote this whole set up where PRs would spin up an environment where we could play with the app. I wrote a bunch of Terraform stuff to start to try to replicate this process. Also, I wanted to get back into the swing of writing Terraform stuff.

TL;DR

• Infrastructure-as-code added with Terraform (VPC, EC2, ECR, RDS, S3, DynamoDB, ACM, IAM, Security Groups).
• Production Docker setup hardened (compose, Nginx reverse proxy, prod Dockerfiles).
• One-command developer ergonomics with a new Makefile.
• Deployment docs and AWS setup scripts added.
• Frontend expanded (Dude Chat page), enhanced navigation, and middleware added.

Infrastructure & DevOps

• Terraform introduced under terraform/ with modular design:
  • Modules for vpc, ec2, ecr, rds, s3, dynamodb, acm, iam, security-groups
  • Environment overlays: terraform/environments/dev/terraform.tfvars and .../prod/...
  • Root wiring and outputs (main.tf, variables.tf, outputs.tf, versions.tf)
  • Deployment notes: terraform/TERRAFORM_DEPLOYMENT.md
• Nginx reverse proxy added for production
  • Config in nginx/nginx.conf to front Next.js and FastAPI behind 80/443
• Docker and Compose hardened
  • docker-compose.prod.yml updated for prod services and health checks
  • Backend prod image split: backend/Dockerfile.prod
  • Repo-level Dockerfile and .dockerignore updated for smaller builds
• One-command local and CI helpers
  • Makefile with targets for dev/test/lint/format/logs/clean
  • scripts/setup-aws-infrastructure.sh to bootstrap AWS resources
  • Env scaffolding: .env.local.backup, env.prod.example, main-example.yml, deploy-example.yml
• Documentation
  • AWS_DEPLOYMENT.md and GITHUB_CONFIGURATION.md added for CI/CD and platform setup
  • Screenshot references and prod compose docs updated

App & Frontend

• Meetings page (previously shipped) was the stepping stone. Since then:
  • New page: src/app/community/dude-chat/page.tsx (community engagement surface)
  • Navigation enhancements in src/components/Navigation.tsx to expose new sections
  • Middleware added (src/middleware.ts) to prep for auth/routing concerns
  • Community detail page edits: src/app/community/[questionId]/page.tsx UI polish

Why this matters

• Repeatable infra: Terraform lets you bring up/down consistent dev/prod stacks.
• Secure, scalable edge: Nginx centralizes TLS and routing, simplifying service topology.
• Faster iteration: Makefile + scripts reduce yak-shaving for common tasks.
• CI/CD readiness: Clear docs + examples accelerate pipeline setup and secrets configuration.
• Product surface growth: Community chat and nav improvements increase engagement and discovery.

What to do next

• Wire Terraform outputs into GitHub Actions variables to fully automate deploys.

• Add health and smoke checks post-deploy (curl API, ping frontend).

• Route logs/metrics to CloudWatch or another aggregator.

• Expand middleware to enforce auth where needed (e.g., chat).

• Gradually swap mock data for real services (DB, queues, analytics events).

• New infra-as-code (terraform/) and prod Nginx/Compose assets added.

• Makefile and scripts streamline local dev and deployment.

• Community features expanded (Dude Chat), navigation and middleware added.

• Docs and env scaffolding round out a CI/CD-ready, production-focused repo.

Written by Kevin Coyle