Projects & architecture

Production-grade cloud and infrastructure builds, from an explorable event-driven platform to a multi-region disaster-recovery blueprint designed to stay online when a whole region fails.

← Back to portfolio
▶ Launch interactive demo
01 · INTERACTIVE DEMO

Event-Driven SaaS Platform on AWS

An interactive, explorable production topology featuring edge security, a multi-AZ VPC of Fargate microservices, an asynchronous event backbone, and cross-cutting security, observability & CI/CD planes. Filter any plane or run a live request trace right in the browser.

  • Edge & Network Security. WAF-fronted edge protecting a multi-AZ VPC that isolates public and private service tiers.
  • Event Backbone. Decoupled Fargate microservices communicating asynchronously through an event-driven messaging layer.
  • Cross-Cutting Planes. Toggle security, observability, and CI/CD overlays to see how each concern threads through the platform.
AWS Fargate Multi-AZ VPC Event-Driven Microservices Observability CI/CD
Launch Interactive Demo ↗
AWS Multi-Region Active-Passive Failover architecture diagram
02 · AWS DISASTER RECOVERY

Multi-Region Active-Passive Failover with Auto-Recovery

A production-grade, event-driven disaster-recovery blueprint designed to survive catastrophic regional failures while preserving client connectivity.

  • Edge Resiliency. Amazon CloudFront Origin Groups route traffic away from a failing origin instantly, without waiting for slow DNS TTL cutovers.
  • Automated State Recovery. CloudWatch monitors 5xx errors, alerts via SNS, and triggers a Lambda function that promotes the standby RDS PostgreSQL replica, with no manual intervention.
  • Graceful Degradation. An S3-hosted static maintenance page catches user traffic during the brief promotion window, preventing transaction timeouts.
Amazon CloudFront RDS PostgreSQL CloudWatch SNS AWS Lambda Amazon S3