Mike Castro
Back to Projects

BankRolls Tools Event-Tracker

High-performance affiliate tracking system

About This Project

Designed and built track.bankrolls.tools from scratch using event-driven architecture, PostgreSQL, Elasticache, Redis, and AWS SQS. Developed a high-performance API capable of handling 300 concurrent requests per second with sub-200ms average response times. Built CI/CD workflows and Dockerized environments with GitHub Actions, supporting zero-downtime deployments, health checks, and rollback mechanisms.

My Contributions

Lead Full Stack Developer

Spike Up Media, LLC

Led architecture and full implementation from scratch as the primary developer

  • Designed and built track.bankrolls.tools from scratch, processing tens of thousands of events daily using event-driven architecture with PostgreSQL, ElastiCache, Redis, AWS SQS, and SNS
  • Developed a high-performance API handling 300 concurrent requests per second with sub-200ms response times
  • Executed 100+ zero-downtime deployments through CI/CD workflows with GitHub Actions, health checks, and rollback mechanisms
  • Rewrote seeders and job workers to batch ElastiCache operations, reducing execution time from 1 minute to near-instant
  • Defined a normalized, polymorphic database schema in collaboration with the CTO, translating evolving business goals into scalable data models
  • Architected a pub/sub model with AWS SNS and systemd-managed consumers for scalable event processing
  • Authored robust infrastructure setup scripts and internal tooling for Redis, SQS, PostgreSQL, S3, and CloudWatch

Screenshots

BankRolls Tools Event-Tracker screenshot 1

Project Details

Role: Senior → Lead DeveloperTeam: 4-6 developers
Production SaaS

Tech Stack

TypeScript

Implemented end-to-end type safety from database models to frontend components

Next.js

Architected and built the entire application from scratch using Next.js 15

Drizzle ORM

Designed normalized database schema, managed migrations, and optimized complex queries for high-volume event processing

PostgreSQL

Designed normalized schema with polymorphic relationships and tuned performance for tens of thousands of daily events

Redis

Built caching layer for API responses and implemented job queues for background processing

ElastiCache

Configured ElastiCache for production caching, rewrote batch operations reducing execution time from 1 minute to near-instant

SQS

Implemented event queuing with decoupled producers/consumers and retry logic with dead-letter handling

SNS

Architected pub/sub model for fan-out messaging and event notifications across services

EC2

Provisioned and configured EC2 instances as primary deployment environment

Cloudflare

Configured DDoS protection and security measures

Jest

Wrote comprehensive unit and integration tests maintaining high test coverage

PM2

Managed Node.js processes in production with auto-restart on failure and worker process management

Systemd

Configured systemd-managed consumers for scalable event processing with auto-restart

GitHub Actions

Built CI/CD workflows supporting 100+ zero-downtime deployments with health checks and rollback mechanisms