Backend Engineer — Kolkata, India
Event-driven services, high-throughput APIs and multi-cloud infrastructure — designed for the day the traffic doubles and nobody is watching.
Three rules I hold to. Everything below them is negotiable.
Queues, retries and idempotency before features. Moving HawkShield's escalation flow onto async messaging removed 60% of the manual incident handling — because the failure path stopped needing a human in it.
Prometheus and Grafana go in before the refactor. The 55% API latency reduction was a measurement problem long before it was a code problem; the profile named the three endpoints that mattered and the rest stayed untouched.
Terraform modules over console clicks. Provisioning became 80% faster and compliance reached 95% across AWS, Azure and GCP — not by working harder, but by making the correct thing the only thing that is easy to do.
Two years of production work, in the only terms that settle an argument.
I am a backend engineer. I spend my days on the unglamorous half of software: the queues, the schemas, the provisioning, the alert that fires at 3am.
Most of my work sits at NextZen Minds, where I build event-driven microservices and multi-cloud infrastructure for security, cloud posture and real-time education platforms. I care about systems that are fast, observable and secure, in that order of difficulty.
The work I am proudest of is invisible: a service that absorbed a traffic multiple without anyone filing a ticket.
Aditya Kumar GuptaThree production systems, and what they measurably changed.
Security · Incident Response
A distributed platform for triaging and escalating security alerts. Async microservices over RabbitMQ, Redis and PostgreSQL, so an alert storm queues instead of collapsing the service behind it.
Cloud · Infrastructure
Multi-cloud infrastructure management with Terraform automation and full observability across AWS, Azure and GCP. One module set, three providers, no console archaeology.
Education · Real-time
A live-learning platform holding a thousand concurrent students: WebSocket class delivery, queued notifications and a PostgreSQL layer rewritten around the queries that were actually slow.
Away from the client work: a product of my own, built and shipped end to end.
Own product · Interview preparation
Practice smarter. Grow faster.
Practivia reads your résumé and the job description you are aiming at, then builds the question set that lives in the gap between the two. Not a generic list of interview questions: the ones that particular pairing invites.
Written answers are graded out of a hundred against hand-curated rubrics that do not drift, so the same answer scores the same next month. The feedback names the points you covered and the ones you missed, which is the part a bare score never tells you.
In active build · Live on Google Play
“How do you stop a queue collapsing under an alert storm?”
0 out of 100
The rubric is fixed. The same answer scores the same next year.
Hover any entry for where it earns its place.
Thirty entries. Each one has shipped.
Where the work was done.
NextZen Minds · Kolkata
Production backend systems across security, cloud and education — from microservice decomposition through to multi-cloud infrastructure and the observability that keeps it honest.
Node.js · NestJS · RabbitMQ · Redis · PostgreSQL · Terraform · AWS / Azure / GCP
CBNITS · Kolkata
First production exposure: REST API development, agile delivery and the deployment pipelines that turn a merge into a running service.
REST APIs · Backend Development · Git · Deployment
B.Tech, Computer Science & Engineering — algorithms, distributed systems, system design.
2019 — 2023Open to backend engineering and distributed systems roles, full-time or remote. A short note about the system and its load is enough to start.
Kolkata, India — working worldwide