Gameplay Learner

Case study

Overview

Served as the Technical Lead for the modernization and productionization of an inherited Next.js-based American football learning platform. The platform helped players learn team-specific gameplays created by coaches and team administrators using a spaced repetition approach. It included a web application for Team Admins and coaches and a dedicated mobile experience for players. Led the technical assessment of the existing codebase, source-code recovery, environment setup, application stabilization, architecture review, AWS deployment, database improvements, containerization, and CI/CD modernization.

Case study

Problem

The client had spent months and years with a previous development team but did not have a stable, deployable product and had only seen the application running on the developers' machines. Recovering the source code itself took considerable effort, followed by extensive dependency, configuration, environment, and build issues that prevented the application from running reliably on a new development environment. The immediate business priority was to get the application hosted on AWS. During the technical review, we also identified architectural concerns around subscription management, where a custom SDK was being used to manage subscription methods and store subscription data in MongoDB despite the highly relational nature of the data and the need for transactional consistency.

Case study

Solution

First recovered and gained control of the inherited source code and resolved the dependency, configuration, environment, and build issues required to make the application reproducible and deployable outside the previous developers' machines. The initial production deployment was established on AWS EC2 to provide the client with a working hosted application. During the architecture review, subscription-related relational data was redesigned and moved from MongoDB to MySQL/Aurora to provide stronger relationships, data integrity, transactional consistency, and ACID guarantees while retaining Stripe for payment processing. The application was subsequently containerized using Docker, automated CI/CD was introduced with GitHub Actions, and the infrastructure was later modernized from EC2 to AWS ECS serverless architecture with AWS CodePipeline in collaboration with the DevOps team.

Architecture

System overview.

The platform consisted of separate Next.js-based frontend and backend applications with dedicated application and data layers. AWS S3 was used for object storage, Aurora RDBMS for relational application and subscription data, SES for email delivery, SNS for notifications, and IAM for access control. Docker was used for application containerization and GitHub Actions supported automated CI/CD during the initial modernization phase. The first production deployment was established on AWS EC2 and was later migrated to AWS ECS serverless architecture with AWS CodePipeline in collaboration with the DevOps team. Stripe handled payment processing while subscription and relational business data were managed through Aurora/MySQL.

Implementation

Key features.

Technical decisions

Challenges & solutions.

Outcome

Measured impact.

• — Converted an inherited application that previously only worked on the original developers' machines into a deployable production system.
• — Established a stable AWS hosting environment and production deployment process.
• — Improved data integrity and transactional reliability by moving subscription-related relational data to Aurora/MySQL.
• — Introduced Docker-based containerization and automated CI/CD for more consistent and repeatable deployments.
• — Modernized the infrastructure from EC2 to AWS ECS serverless architecture for improved scalability and operational flexibility.
• — Established a stronger technical foundation for continued product development and future scaling.