I build backend systems where reliability isn't optional. My focus is distributed systems in Node.js and Python real-time event pipelines, concurrent reservation engines, and fault-tolerant async processors. I care about data durability, observable systems, and writing code that behaves correctly under load, not just in development.
Currently looking for a backend or full-stack role where system design decisions matter.
Production-grade reservation system for high-traffic flash sales.
- Eliminated race conditions with Redis Lua atomic inventory checks (no TOCTOU window)
- Idempotent BullMQ order fulfillment duplicate Stripe webhooks never double-process
- Concurrency integration test: 6 simultaneous requests, inventory of 3, exactly 3 succeed
Distributed real-time analytics pipeline with a fault-tolerant stream worker.
- Redis Stream consumer with crash-safe bookmark resumes exactly on restart
- TimescaleDB time_bucket upserts + live Socket.IO dashboard updates
- 80%+ test coverage: unit, integration (Docker), and full E2E pipeline assertion
Resilient webhook processor using Redis LPUSH/BRPOP queue.
- Automatic retry, dead-letter tracking, and idempotent handlers
- CLI tool for queue inspection and manual intervention
- Admin dashboard for real-time job monitoring
Production-ready FastAPI authentication microservice.
- JWT access tokens + opaque refresh tokens hashed in database
- Argon2 password hashing, Google OAuth, and email verification via Resend
- SOLID architecture with abstract base classes and Alembic migrations
- Advanced System Design - Distributed systems and microservices patterns
- Database Optimization - Query performance and indexing strategies
- Message Queue Systems - BullMQ, Redis Streams, and event-driven architecture
- Production Best Practices - Monitoring, observability, and failure recovery
I'm always open to discussing technology, collaboration opportunities, or just geeking out about system design!
⭐ Feel free to explore my repositories and reach out if you'd like to collaborate!