Open to opportunities

Abhishek Kumar

@abhishekkumar75

Message

Backend SDE intern building high-performance distributed systems with Node.js, Redis, and pgBouncer.

India

Message

What I'm looking for

I’m looking for a backend/distributed-systems role where I can design reliable Node.js services, push latency down with load/chaos testing, and build robust rate limiting, job queues, and connection-safe infrastructure.

I’m a CS BTech graduate and backend-focused SDE intern who builds core systems for high-traffic platforms—where reliability and measured performance matter as much as features.

At Webedigital, I architected a two-process Node.js backend (Express/Socket.IO API plus a BullMQ worker tier) with Redis-backed job queues and pgBouncer connection pooling. I implemented Redis Lua rate limiting, sustained 99.90% success across 63,692 requests at 300 VUs, and cut p95 latency from 1.97s to 141ms by fixing bottlenecks through caching and composite indexing.

I also strengthened correctness under pressure by writing multiple k6 load test suites and diagnosing pool exhaustion at 200 VUs. Beyond performance, I designed a trust and moderation system with event-driven score transitions, device-level ban enforcement, and append-only audit logging.

Earlier at Doctly, I built an end-to-end doctor verification workflow with a 3-state approval flow, admin dashboard, and Nodemailer email notifications, then delivered MERN stack APIs for search, blog CMS, and profile management. My projects—like a distributed webhook delivery engine with at-least-once delivery and chaos testing, and a horizontally scalable URL shortener—reflect my drive to engineer resilient distributed systems.

Experience

Work history, roles, and key accomplishments

Current

Backend SDE Intern

Current

Webedigital

Mar 2026 - Present (4 months)

Architected a two-process Node.js backend with an Express/Socket.IO API service and a BullMQ worker tier, using Redis-backed job queues and pgBouncer connection pooling. Implemented Redis Lua rate limiting and load-tested with k6, achieving 99.90% success over 63,692 requests at 300 VUs and reducing p95 latency from 1.97s to 141ms.

Socket.IO BullMQ Redis PG Bouncer Rate Limits Caching Audit Logging

MERN Stack Developer Intern

Doctly

Jul 2025 - Dec 2025 (5 months)

Built a doctor verification workflow with a 3-state approval flow, admin review dashboard, and Nodemailer email notifications on state transitions. Developed paginated doctor search, blog CMS, and profile management APIs across the MERN stack, including a 16-task integration dry run ahead of release.

MERN Stack MongoDB Nodemailer REST APIs Pagination CMS Notifications Integration Testing

Distributed Webhook Delivery Engine

GitHub

Jun 2026 - Present (1 month)

Built a distributed webhook delivery engine with at-least-once delivery using Redis Lua-scripted distributed locking and two-layer idempotency (Redis fast-path plus PostgreSQL ON CONFLICT). Added jittered exponential backoff, dead-letter queue, outbox reconciler, circuit breaker, and Redis Lua token-bucket rate limiting; validated with Dockerode chaos tests and k6 load testing (0 event loss).

BullMQ Redis PostgreSQL Distributed Locking Dead Letter Queues Outbox Pattern Token Bucket Rate Limiting Dockerode Chaos Testing

URL Shortener – Distributed System

GitHub

Jan 2026 - Present (6 months)

Designed a horizontally scalable URL shortener using consistent hashing and Redis caching to avoid rebalancing during node scaling. Load- and failure-tested with k6 to identify saturation limits and maintain graceful degradation during partial outages.

Redis PostgreSQL Consistent Hashing Caching Load Testing Failure Testing Graceful Degradation