My AWS Glow-Up Story Powered by League of Legends

I'd used AWS before, spinning up EC2 instances, basic S3 storage, the usual stuff. But I'd never built a complete application with proper architecture and modern deployment practices. The Riot Games Rift Rewind Developer Challenge changed that by walking through building something real instead of just isolated services.

Having an AWS account already meant I could skip the basic setup, but the challenge's security-first approach was still valuable. The emphasis on private S3 buckets by default, proper IAM roles, and encrypted parameter storage reinforced good habits I'd been sloppy about before.

The IAM role setup for Lambda to access Systems Manager Parameter Store was cleaner than hardcoding API keys, which I'll admit I'd done in quick projects before. Seeing the proper way laid out step-by-step was useful.

I'd used CloudFront before but never really understood the why behind it. The challenge explained the distance problem clearly, users in Tokyo waiting for data from Mexico City servers. That 22,000 kilometer round trip explanation made CDNs click in a way they hadn't before.

Setting up the origin access control to keep S3 private while serving through CloudFront was straightforward once you see the pattern. Much better than making buckets public, which I'd definitely done before for quick demos.

This was the biggest upgrade to my process. I'd been manually uploading files to S3 and dealing with cache invalidations like some kind of caveman. The GitHub + Amplify integration eliminated all that manual work.

Push code, automatic deployment, global distribution, HTTPS - all handled automatically. This is apparently how professional teams work, and now I get why. The productivity difference is huge once you stop doing deployments manually.

The Lambda function for Riot API calls was a good example of serverless done right. Function only runs when needed, scales automatically, costs almost nothing. I'd used Lambda before but usually for scheduled tasks, not as an API backend.

The CORS configuration and Function URL setup was cleaner than setting up API Gateway, which I'd struggled with in previous projects. Sometimes the simpler approach is better.

Storing the Riot API key in Parameter Store instead of environment variables was another good practice. More secure and easier to rotate keys when needed.

The challenge connected services I'd used separately into a complete application. S3, CloudFront, Lambda, GitHub integration - I'd touched all these before but never as a cohesive system.

The automated deployment pipeline was the biggest win. Being able to iterate quickly without manual deployment steps changes how you approach development. You actually want to make small improvements when deploying them is effortless.

The security practices throughout were solid reinforcement. Private by default, proper IAM roles, encrypted secrets. Good habits that are easy to skip when you're moving fast.

By the end, I had a project running on infrastructure that scales globally and costs practically nothing to operate. CloudFront handles the global distribution, Lambda runs the API calls only when needed, and the whole thing updates automatically when I push code changes.

The architecture is clean - static assets served globally, API calls handled serverless, proper security throughout. It's the kind of setup I'd actually use for real projects now.

Even with existing AWS experience, following the challenge structure was valuable. It connected scattered knowledge into a complete workflow and reinforced security practices I'd been inconsistent about.

The progression from basic storage to global distribution to automated deployment to API integration made sense. Each day built on the previous without feeling repetitive.

The project is live at Live Site  and the code is on GitHub at github-repo . Good example of modern cloud architecture done right.