Stack: Node.js, Express, MongoDB, Redis, Cloudinary , Zod, Docker, Swagger
A scalable and production-grade backend designed for e-commerce applications. Implemented role-based architecture, token rotation, and IP-based rate limiting to ensure secure and efficient access control. Reused the authentication API for token management to simplify reusability and maintainability across different modules.
Integrated cart management using Redis for high-performance caching and session persistence, ensuring minimal latency for real-time cart updates. Added API documentation with Swagger for smooth developer onboarding and testing. Includes Zod-based schema validation, Dockerized container setup, and environment-specific configurations for scalability.
Focus: Security • Scalability • Reusability • Clean Architecture
Stack: Node.js, Express, Redis, Docker, Jest
ReTraced is a transparent and extensible distributed job scheduler built to make retry behavior, failure handling, and job lifecycle transitions explicit and observable.
ReTraced is designed in such a way that every job can define its own retry strategy. Although not every job requires a unique strategy, different classes of failures (network, infra, logic, business rules) demand different retry behaviors.
Focus: Security • Scalability • Reusability • Clean Architecture
Stack: TypeScript, Node.js, Express, Redis, Lua, Docker, Zod
Built a production-grade distributed rate limiter supporting Leaky Bucket, Fixed Window, Sliding Window, and Token Bucket algorithms, using Lua scripts for atomic Redis operations to ensure correctness under concurrent access.
Deployed a horizontally scalable Redis cluster with 3 masters and 3 replicas, ensuring race-free consistency and fault tolerance under sustained peak traffic.
Focus: Concurrency • Scalability • Fault Tolerance • Backend Systems