I build distributed, production-grade systems - not just web apps. Currently focused on cloud-native backend engineering: event-driven architectures, container orchestration, and systems that survive real-world failure conditions.
📍 Islamabad, Pakistan · BS-IT @ University of Chakwal (Jun 2026) 🏅 NSCT'26 — Top 3.3% Nationally · Ranked 96.7th percentile in the National Skill Competency Test by HEC, PSEB & P@SHA
Sentinel-SRE · Distributed Uptime Monitor
Production-grade distributed system monitoring website health at scale.
- Zero data loss under worker failure — RabbitMQ Manual ACKs + Dead Letter Queues + Exponential Backoff
- Sub-millisecond status lookups — Redis caching layer cuts PostgreSQL read I/O dramatically
- Full observability — Prometheus + Grafana dashboards tracking P99 latency per endpoint
- Load verified — k6 stress tested at concurrent user load to identify breaking points before production
Go Node.js RabbitMQ Redis PostgreSQL Prometheus Grafana Docker k6
Pulse_Queue · production-grade idempotent job queue built with Node.js, Express, PostgreSQL, and Redis.
- Exactly-once processing — PostgreSQL advisory locks let workers claim jobs non-blockingly, so no two workers ever process the same job concurrently
- Request-level idempotency — SHA-256 request fingerprinting + 24-hour Redis TTL cache returns the original response to retried requests without re-executing side effects
- Self-healing job recovery — a stale
updated_aton a job stuck in "processing" signals a crashed worker, and the scheduler resets it to pending automatically - Scalable pagination — cursor-based
(created_at, id)navigation backed by a matching composite index keeps query performance constant regardless of page depth
Node.js Express PostgreSQL Redis Docker worker_threads pg_try_advisory_lock
Link Vault · Kubernetes-Orchestrated Asset Manager
DevOps-first MERN application built for production-grade Kubernetes deployment.
- Self-healing infrastructure — Kubernetes automated pod recovery and restart policies
- Config decoupling — Kubernetes Secrets + ConfigMaps for clean credential isolation
- Optimized images — multi-stage Dockerfiles for lightweight production containers
- Live monitoring — Prometheus + cAdvisor + Grafana dashboards for real-time resource visibility
React Node.js MongoDB Docker Kubernetes Prometheus Grafana cAdvisor
Languages Go · JavaScript/TypeScript · Python
Backend Node.js · Express.js · FastAPI
Frontend React · Vite · TailwindCSS
Databases PostgreSQL · MongoDB · Redis · MySQL
Messaging RabbitMQ · BullMQ
Infra Docker · Kubernetes · GitHub Actions
Observability Prometheus · Grafana · k6
-
Open Source SigNoz contributor — feat(components): add Skeleton (PR #230) +712 lines · 18 review comments · merged May 2026 github.com/SigNoz/components/pull/230
-
NSCT'26 — Top 3.3% nationally (96.7th percentile) — National Skill Competency Test by HEC, PSEB & P@SHA
-
Dev Connect Hub — Founded a 100+ member developer community, university-recognized, Islamabad
858x Faster: What I Learned Optimising 500,000 PostgreSQL Rows https://www.linkedin.com/posts/afshan-qasim-998917300_postgresql-databaseperformance-queryoptimization-share-7463599947624611841-L436/?utm_source=social_share_send&utm_medium=member_desktop_web&rcm=ACoAAE0FscQB0wnD66Ar3MdwCW-l06nUW95fayw
Your index is not a fix. It is a ticking clock. https://www.linkedin.com/posts/afshan-qasim-998917300_postgresql-backendengineering-sre-share-7464246699239239680-D-1G/?utm_source=social_share_send&utm_medium=member_desktop_web&rcm=ACoAAE0FscQB0wnD66Ar3MdwCW-l06nUW95fayw
1,820x Faster: What I Learned Optimizing 500,000 PostgreSQL Rows Read the full article on dev.to
- Contributing to open source - SigNoz observability platform (React · Go)
- Deepening Kubernetes operator patterns and Go concurrency internals