Backend engineer at SDE-3 level with 5+ years building REST APIs, event-driven services and real-time systems in Node.js , TypeScript , NestJS and Express.js . Owns backend work end to end: PostgreSQL schema design and zero-downtime migrations, Kafka and RabbitMQ event pipelines, Redis caching, Elasticsearch search, Socket.IO WebSocket gateways, and production monitoring with Prometheus and Grafana . Has run these systems in EdTech, WhatsApp messaging automation and multiplayer gaming, with measured results including 40% lower search latency , 35% faster API responses and 99.9% uptime at ~100K messages per bot per day . Takes feature scope and high-level design decisions directly with clients, documents the approach before starting a task so scope, effort and trade-offs are agreed in writing, and proves the design out as a POC before committing to a build. Works comfortably as the only backend engineer on a service or inside a team.
Gurugram, HR
- Backend ownership: sole owner and technical point of contact for a production multiplayer learning game backend, built as a modular NestJS service with DTOs, guards, interceptors and a separated service layer, so several developers could ship features in parallel without merge conflicts.
- Scope and technical decisions: works directly with clients and product on feature scope and the high-level design calls behind it, then turns those into backend deliverables, API contracts and a delivery order the team can pick up.
- Proof of concept work: builds POCs for new features and integrations so the approach, cost and failure modes are settled before full implementation begins.
- Search service: indexed ~12 million dictionary entries in an Elasticsearch cluster with Redis caching in front, cutting average lookup latency by 40% and keeping results consistent under high concurrent load.
-
Real-time WebSocket layer:
built the Socket.IO gateway for live game-state synchronisation, player matchmaking and multiplayer
feedback, with a consistent
${eventName}:errorevent contract so client teams could detect and handle failures without guesswork. - Queues and workers: moved power-up processing (Magic Word, Highlight) into BullMQ background jobs, taking the work off the request path and cutting average API response time by ~35% across active game sessions.
- PostgreSQL schema and migrations: designed the learning module schema covering spaced repetition, difficulty progression and learner progress tracking, and ran the full TypeORM migration lifecycle: adding NOT NULL columns as nullable, backfilling, then tightening the constraint to avoid downtime.
- Monitoring and deployment: instrumented services with Prometheus metrics and Grafana dashboards for health and traffic, cutting mean time to detect incidents by 50% ; every service runs in Docker with per-environment configuration.
- API security: implemented JWT authentication, rate limiting, field-level encryption and NSFW content moderation, plus reusable webhook handlers and third-party integration modules that other teams pick up directly.
- Custom Emailer Service: built a standalone email delivery tool for the internal dashboard of a renowned client, with custom To/CC configuration, dynamic attachments, per-report settings and scheduled delivery. It now carries the daily grievance reporting workflow for 20 to 25 management recipients.
- Delivery: runs sprint planning, writes architecture documentation with Mermaid.js diagrams, manages Jira workflows and reviews pull requests to keep standards consistent across the backend team.
Gurugram, HR
- INTERCOMM platform backend: built and scaled a WhatsApp Business API platform as the only backend engineer, covering Campaign Builder, Template Creator, Custom Flow Designer and analytics. It processed 1M+ campaign messages per quarter , contributed to a 10× expansion in business reach , doubled sales revenue and a measured 45% improvement in campaign ROI .
- Queue and webhook throughput: sustained ~100,000 messages per day per bot using Redis-backed queue management, webhook event processing and async job handling, holding 99.9% uptime under sustained load.
- BridgeQ event-driven backbone: architected a multi-channel messaging service on Kafka for async event communication and RabbitMQ for event routing, decoupling services and improving resilience. A provider-credential abstraction layer let new messaging channels be added without touching core business logic.
- REST APIs and monitoring: shipped production REST APIs and internal dashboards instrumented with Prometheus and Grafana, and integrated third-party provider APIs behind webhook and adapter patterns new engineers could follow on their own.
- CI/CD and documentation: owned the GitHub Actions pipelines, set the Git workflow standards the team uses, and wrote the technical documentation and runbooks still in active use.
Chandigarh
- Notification scheduling service: designed and shipped a data-driven push notification scheduler in Node.js and Express with personalised payloads and send-time selection, which measurably improved user retention across the platform.
- Database administration and tuning: administered AWS RDS and Apache Cassandra instances and applied SQL performance tuning across schemas, indexes and query patterns to keep notification delivery fast and available at scale.
Chandigarh
- Course content APIs: built and maintained the RESTful APIs behind the Website Plan Data module, serving segmented course content to thousands of students every day.
- KYS backend: sole developer on Know Your Student, delivering encrypted REST APIs whose pop-up and form logic came entirely from database-table configuration, so non-technical teams could change behaviour without a code deployment.
- Dynamic forms and data integration: built dynamic form management, real-time metadata updates and AWS RDS integration inside an Agile team, with code review and documentation habits carried into later roles.
- Designed and maintained the whole backend for a production multiplayer platform: real-time gameplay over WebSockets, ELO-based matchmaking and a multi-tenant admin dashboard.
- Indexed ~12 million dictionary entries across Elasticsearch and Redis, sustaining 200K+ concurrent player sessions with 40% lower lookup latency than the previous implementation.
- Moved power-up processing to BullMQ workers, cutting request-path API response times by ~35% , and modelled spaced repetition, difficulty progression and learner progress in PostgreSQL.
- Added Prometheus and Grafana instrumentation across services, reducing incident detection time by 50% in production.
- Sole backend engineer for Campaign Builder, Template Creator, Custom Flow Designer and analytics, together processing 1M+ messages per quarter .
- Built the event-driven messaging backbone on Kafka and RabbitMQ, holding ~100K messages/day/bot at 99.9% uptime with a 45% improvement in campaign ROI .
- Supported a 10× expansion in client business reach and doubled platform sales revenue within the first year of launch.
- Event-driven architecture (Kafka + RabbitMQ) unified SMS, Email and FCM push under a single routing layer for a hyperlocal QSR platform prototype.
- Provider-credential abstraction reduced new-channel integration to configuration only; zero core logic changes required.
- Batched and scheduled delivery for a renowned client, built to keep a recurring dispatch on time without a person driving it.
- Custom To/CC configs, dynamic attachments and per-report delivery profiles for daily grievance and management reporting.
- Sole developer on the notification scheduling infrastructure, building personalised payload logic and optimal send-time selection that measurably improved retention on a large-scale EdTech platform.
- Tuned SQL on AWS RDS and optimised Cassandra query patterns to keep notification delivery highly available at scale.
- Built the KYS backend solo: encrypted REST APIs driving pop-up and form behaviour from database-table configuration, giving non-technical teams control over the student experience without deployments.
- Delivered real-time metadata updates and form interactivity that improved student onboarding conversion on the PrepLadder platform.
- Creator and maintainer of Numsy, an open-source library, CLI and local dashboard that validates, extracts and de-duplicates Indian mobile numbers from CSV, Excel and text, published on npm as numsy.
- Built one browser-safe engine shared by the library, the CLI, a NestJS dashboard on localhost and a website playground that runs in a Web Worker, so every surface returns the same result.
- Validates about 1.1 million numbers per second and analyses and de-duplicates 500,000 rows in under 2 s, verified on real workbooks of up to 505,277 rows.
- Gated by 145 unit and API tests and Playwright end-to-end suites that check every dashboard figure against the engine, with releases cut by semantic-release.