
Full-Stack E-Commerce Platform
> Modern E-Commerce Storefront & REST API with Dynamic Multi-Image SEO, Automated CI/CD & Terraform IaC
A production-grade full-stack e-commerce platform built from scratch with a decoupled architecture pairing a responsive React storefront with an Express/Node.js REST API. Designed for high conversion and discoverability, the platform features dynamic multi-image galleries engineered for fast Largest Contentful Paint (LCP) and SEO, persistent cart & wishlist synchronization, comprehensive product catalog filtering, interactive Swagger API documentation, automated two-stage GitHub Actions CI/CD workflows, and AWS cloud infrastructure declared in Terraform.
Decoupled full-stack architecture with a React SPA client communicating over RESTful JSON with an Express/Node.js backend, backed by MongoDB for flexible catalog schemas, Docker for consistent containerization, GitHub Actions for automated pipelines, and Terraform for AWS infrastructure.
Seamless customer purchasing experience with real-time catalog search, multi-angle product image galleries, and persistent cart synchronization.
Administrative control over multi-image product uploads, category taxonomy, inventory stock counts, and order status transitions.
Continuous Integration and Deployment pipelines connected to version-controlled AWS cloud infrastructure.
Product detail pages featuring multiple high-resolution image angles caused slow Largest Contentful Paint (LCP) and high client bandwidth consumption.
Engineered dynamic multi-image carousels with responsive image breakpoints, priority preloading for the primary hero shot, and deferred thumbnail loading, maintaining sub-second render times and optimal SEO indexability.
Manual deployment and inconsistent local runtime environments risked shipping breaking regressions and configuration drift.
Built separate CI (Continuous Integration: automated lint, test, build validation) and CD (Continuous Deployment: Docker image creation and deployment orchestration) workflows in GitHub Actions.
Manually configuring cloud resources on AWS is error-prone, hard to audit, and challenging to replicate across staging and production environments.
Authored modular Terraform infrastructure scripts declaring AWS S3 buckets for product image assets, bucket access policies, and environment configurations in version-controlled code.