LeetGrammar
A Somali grammar app with lessons and practice. Its planned AWS migration brings infrastructure into Terraform and GitHub Actions.
AWS migration planned
Target architecture
- Target: Terraform-defined infrastructure
- Target: ALB + Auto Scaling across 2 AZs
- Target: accounts and durable progress
- Target: delivery through GitHub Actions
EC2 · Auto Scaling · ALB · RDS · Cognito · CloudWatch · Terraform · GitHub Actions
Planned architecture
Scroll the diagram horizontally to explore the full flow.
The product
LeetGrammar is a language-learning application I built to turn structured practice into a focused daily workflow. The product includes interactive exercises, progress state, and a responsive interface designed around short study sessions.
Why the infrastructure is changing
Managed hosting made the first version fast to ship, but it also hid instance lifecycle, identity, persistence, and observability. The next version moves those responsibilities into infrastructure I can inspect and operate.
The target architecture uses an Application Load Balancer and Auto Scaling across two Availability Zones, RDS for durable data, Cognito for accounts, and CloudWatch for metrics and alarms.
Infrastructure as code
The planned delivery workflow uses Terraform to define the stack and GitHub Actions to apply it. The goal is for the repository—not a collection of console screenshots—to be the reproducible artifact.
Current state
The application exists while the AWS migration remains planned. I present those as separate facts: the software exists, and the new infrastructure is a target, not a completed deployment.
- React and Vite application
- Responsive learning interface
- Planned Terraform-defined AWS target architecture
- Planned automated delivery workflow
- Current progress stored locally; durable accounts and progress are planned